{
"kind": "page",
"html": "https://insidethemachine.ai/shop",
"md": "https://insidethemachine.ai/shop.md",
"title": "The free shop",
"url": "https://insidethemachine.ai/shop",
"fields": {},
"intro": [
"Data as of 2026-10-26.",
"152 free AI models and 264 open-source tools. The page starts with only the parts that fit the reader's memory (the \"Fits my computer\" switch). Add parts to My setup, then check out. What fits each memory size, in full: https://insidethemachine.ai/index.md. Skill packs ranked by stars gained: https://insidethemachine.ai/skills.md."
],
"sections": [
{
"heading": "Our pick per memory size",
"level": 2,
"text": [
"The smartest model that fits and runs with Ollama on Mac, Windows and Linux. The chat setup sheets install it; the 24/7 helper below may use a smaller brain so the ears and voice fit next to it."
],
"items": [
"8 GB: Qwen3.5 4B, 23% of Claude Opus 5.5, needs 5.8 GB",
"16 GB: Gemma 4 12B, 25% of Claude Opus 5.5, needs 11.2 GB",
"24 GB: Gemma 4 12B, 25% of Claude Opus 5.5, needs 11.2 GB",
"32 GB: Qwen3.8 27B, 59% of Claude Opus 5.5, needs 22.7 GB",
"48 GB: Qwen3.8 27B, 59% of Claude Opus 5.5, needs 22.7 GB",
"64 GB: Qwen3.8 27B, 59% of Claude Opus 5.5, needs 22.7 GB",
"96 GB: Qwen3.8 27B, 59% of Claude Opus 5.5, needs 22.7 GB",
"128 GB: Qwen3.8 27B, 59% of Claude Opus 5.5, needs 22.7 GB",
"192 GB: Qwen3.8-Flash-Next (125B-A6B), 69% of Claude Opus 5.5, needs 140 GB",
"256 GB: Qwen3.8-Flash-Next (125B-A6B), 69% of Claude Opus 5.5, needs 140 GB",
"512 GB: Qwen3.8-Flash-Next (125B-A6B), 69% of Claude Opus 5.5, needs 140 GB"
],
"tables": []
},
{
"heading": "Your 24/7 helper (FIG. 2a)",
"level": 2,
"text": [
"Rule: The brain, the ears (speech to text) and the voice (text to speech) stay loaded together. Ears and voice get up to 15% of the memory for AI each (at least 1 GB). The brain is the best model that can use tools, fits in the rest and runs with Ollama on Mac, Windows and Linux. To make an image or a video, the helper pauses the brain (ollama stop) and that model uses its space.",
"Set it up: 1. A private ChatGPT (https://insidethemachine.ai/do/chat) installs Ollama, the brain and a chat app. 2. Speech to text (https://insidethemachine.ai/do/voice) adds the ears. It installs Whisper, a different speech-to-text model from the one on the map; both turn speech into text. 3. Voice tools (https://insidethemachine.ai/setups/c/voice) list the text-to-speech apps for the voice. We have no recipe for it yet. 4. An agent that does tasks (https://insidethemachine.ai/do/agent) adds an always-on agent app from 32 GB. On 16 GB, use the chat app from step 1. (Shown for 16 GB, the page's default size.)"
],
"items": [
"Note, 8 GB: No video model in our list fits yet. The first memory size where one fits next to the ears and voice is 32 GB.",
"Note, 16 GB: No video model in our list fits yet. The first memory size where one fits next to the ears and voice is 32 GB.",
"Note, 24 GB: No video model in our list fits yet. The first memory size where one fits next to the ears and voice is 32 GB."
],
"tables": [
{
"columns": [
"memory",
"brain (always on)",
"ears",
"voice",
"always on",
"spare",
"image (on demand)",
"video (on demand)"
],
"rows": [
{
"memory": "8 GB",
"brain (always on)": "Qwen3.5 2B 4.2 GB, 12% of Claude Opus 5.5, ollama run qwen3.5:2b",
"ears": "Qwen3-ASR 0.6B 1 GB",
"voice": "Kokoro 82M v1.0 0.5 GB",
"always on": "5.7 of 6 GB",
"spare": "0.3 GB",
"image (on demand)": "Z-Image Turbo (4 GB), runs in stable-diffusion.cpp, ComfyUI or InvokeAI. Data note: 4 GB with low-bit GGUF in stable-diffusion.cpp; maker says full model fits 16 GB.",
"video (on demand)": "none fits (no video model in our list fits yet)"
},
{
"memory": "16 GB",
"brain (always on)": "Qwen3.5 9B 9.6 GB, 23% of Claude Opus 5.5, ollama run qwen3.5:9b",
"ears": "Qwen3-ASR 0.6B 1 GB",
"voice": "Kokoro 82M v1.0 0.5 GB",
"always on": "11.1 of 12 GB",
"spare": "0.9 GB",
"image (on demand)": "Z-Image Turbo (4 GB), runs in stable-diffusion.cpp, ComfyUI or InvokeAI. Data note: 4 GB with low-bit GGUF in stable-diffusion.cpp; maker says full model fits 16 GB.",
"video (on demand)": "none fits (no video model in our list fits yet)"
},
{
"memory": "24 GB",
"brain (always on)": "Gemma 4 12B 11.2 GB, 25% of Claude Opus 5.5, ollama run gemma4:12b",
"ears": "MOSS-Transcribe-Diarize 0.9B 2.5 GB",
"voice": "Magpie TTS Multilingual 357M 2 GB",
"always on": "15.7 of 18 GB",
"spare": "2.3 GB",
"image (on demand)": "Z-Image Turbo (4 GB), runs in stable-diffusion.cpp, ComfyUI or InvokeAI. Data note: 4 GB with low-bit GGUF in stable-diffusion.cpp; maker says full model fits 16 GB.",
"video (on demand)": "none fits (no video model in our list fits yet)"
},
{
"memory": "32 GB",
"brain (always on)": "Gemma 4 12B 11.2 GB, 25% of Claude Opus 5.5, ollama run gemma4:12b",
"ears": "Qwen3-ASR 1.7B 3 GB",
"voice": "Magpie TTS Multilingual 357M 2 GB",
"always on": "16.2 of 24 GB",
"spare": "7.8 GB",
"image (on demand)": "Z-Image Turbo (4 GB), runs in stable-diffusion.cpp, ComfyUI or InvokeAI. Data note: 4 GB with low-bit GGUF in stable-diffusion.cpp; maker says full model fits 16 GB.",
"video (on demand)": "HunyuanVideo 1.5 (14 GB), runs in ComfyUI or stable-diffusion.cpp. Data note: 14 GB GPU with offloading; license excludes the EU, UK and South Korea."
},
{
"memory": "48 GB",
"brain (always on)": "Qwen3.8 27B 22.7 GB, 59% of Claude Opus 5.5, ollama run qwen3.8:27b",
"ears": "Qwen3-ASR 1.7B 3 GB",
"voice": "Magpie TTS Multilingual 357M 2 GB",
"always on": "27.7 of 36 GB",
"spare": "8.3 GB",
"image (on demand)": "FLUX.2 [dev] (24 GB), runs in ComfyUI, stable-diffusion.cpp or InvokeAI. Data note: 24 GB is the maker's 4-bit example on an RTX 4090 with a remote text encoder.",
"video (on demand)": "HunyuanVideo 1.5 (14 GB), runs in ComfyUI or stable-diffusion.cpp. Data note: 14 GB GPU with offloading; license excludes the EU, UK and South Korea."
},
{
"memory": "64 GB",
"brain (always on)": "Qwen3.8 27B 22.7 GB, 59% of Claude Opus 5.5, ollama run qwen3.8:27b",
"ears": "Granite Speech 5.0 2B 6 GB",
"voice": "Magpie TTS Multilingual 357M 2 GB",
"always on": "30.7 of 48 GB",
"spare": "17.3 GB",
"image (on demand)": "FLUX.2 [dev] (24 GB), runs in ComfyUI, stable-diffusion.cpp or InvokeAI. Data note: 24 GB is the maker's 4-bit example on an RTX 4090 with a remote text encoder.",
"video (on demand)": "HunyuanVideo 1.5 (14 GB), runs in ComfyUI or stable-diffusion.cpp. Data note: 14 GB GPU with offloading; license excludes the EU, UK and South Korea."
},
{
"memory": "96 GB",
"brain (always on)": "Qwen3.8 27B 22.7 GB, 59% of Claude Opus 5.5, ollama run qwen3.8:27b",
"ears": "Granite Speech 5.0 2B 6 GB",
"voice": "Breeze TTS 2 8 GB",
"always on": "36.7 of 72 GB",
"spare": "35.3 GB",
"image (on demand)": "Qwen-Image 2.1 (46 GB), runs in stable-diffusion.cpp, ComfyUI or mflux (Mac only). Data note: Memory is mflux bf16 peak (64 GB Mac comfortable, -q 8 for less); research-only license.",
"video (on demand)": "MiniMax H3 (42.5 GB), runs in ComfyUI, stable-diffusion.cpp or InvokeAI. Data note: 42.5 GB is ComfyUI's smallest variant (123.6 GB full); also tops open models on AA text-to-video at 1159."
},
{
"memory": "128 GB",
"brain (always on)": "Qwen3.8 27B 22.7 GB, 59% of Claude Opus 5.5, ollama run qwen3.8:27b",
"ears": "Granite Speech 5.0 2B 6 GB",
"voice": "Breeze TTS 2 8 GB",
"always on": "36.7 of 96 GB",
"spare": "59.3 GB",
"image (on demand)": "Qwen-Image 2.1 (46 GB), runs in stable-diffusion.cpp, ComfyUI or mflux (Mac only). Data note: Memory is mflux bf16 peak (64 GB Mac comfortable, -q 8 for less); research-only license.",
"video (on demand)": "MiniMax H3 (42.5 GB), runs in ComfyUI, stable-diffusion.cpp or InvokeAI. Data note: 42.5 GB is ComfyUI's smallest variant (123.6 GB full); also tops open models on AA text-to-video at 1159."
},
{
"memory": "192 GB",
"brain (always on)": "Qwen3.8 27B 22.7 GB, 59% of Claude Opus 5.5, ollama run qwen3.8:27b",
"ears": "Granite Speech 5.0 2B 6 GB",
"voice": "Breeze TTS 2 8 GB",
"always on": "36.7 of 144 GB",
"spare": "107.3 GB",
"image (on demand)": "Qwen-Image 2.1 (46 GB), runs in stable-diffusion.cpp, ComfyUI or mflux (Mac only). Data note: Memory is mflux bf16 peak (64 GB Mac comfortable, -q 8 for less); research-only license.",
"video (on demand)": "MiniMax H3 (42.5 GB), runs in ComfyUI, stable-diffusion.cpp or InvokeAI. Data note: 42.5 GB is ComfyUI's smallest variant (123.6 GB full); also tops open models on AA text-to-video at 1159."
},
{
"memory": "256 GB",
"brain (always on)": "Qwen3.8-Flash-Next (125B-A6B) 140 GB, 69% of Claude Opus 5.5, ollama run qwen3.8-flash-next:125b-a6b",
"ears": "Granite Speech 5.0 2B 6 GB",
"voice": "Fish Audio Drama 3 (preview) 28 GB",
"always on": "174 of 192 GB",
"spare": "18 GB",
"image (on demand)": "Qwen-Image 2.1 (46 GB), runs in stable-diffusion.cpp, ComfyUI or mflux (Mac only). Data note: Memory is mflux bf16 peak (64 GB Mac comfortable, -q 8 for less); research-only license.",
"video (on demand)": "MiniMax H3 (42.5 GB), runs in ComfyUI, stable-diffusion.cpp or InvokeAI. Data note: 42.5 GB is ComfyUI's smallest variant (123.6 GB full); also tops open models on AA text-to-video at 1159."
},
{
"memory": "512 GB",
"brain (always on)": "Qwen3.8-Flash-Next (125B-A6B) 140 GB, 69% of Claude Opus 5.5, ollama run qwen3.8-flash-next:125b-a6b",
"ears": "Granite Speech 5.0 2B 6 GB",
"voice": "Fish Audio Drama 3 (preview) 28 GB",
"always on": "174 of 384 GB",
"spare": "210 GB",
"image (on demand)": "Qwen-Image 2.1 (46 GB), runs in stable-diffusion.cpp, ComfyUI or mflux (Mac only). Data note: Memory is mflux bf16 peak (64 GB Mac comfortable, -q 8 for less); research-only license.",
"video (on demand)": "MiniMax H3 (42.5 GB), runs in ComfyUI, stable-diffusion.cpp or InvokeAI. Data note: 42.5 GB is ComfyUI's smallest variant (123.6 GB full); also tops open models on AA text-to-video at 1159."
}
]
}
]
},
{
"heading": "The shop: free AI parts, picked for your computer",
"level": 2,
"text": [
"Every part here is free to download and runs on your own computer. Pick an aisle, add parts to \"My setup\" (saved in your browser), then check out right in the My setup panel: your computer (system and memory), how you use it (one way, the chat window picked) and how you run it (fully local, or Claude conducts). One email then gives one folder to download and one prompt for Claude Code or Codex that sets up every part. For an agent, the same answers as JSON: https://insidethemachine.ai/api/checkout?packs=&mode=&way=&os=&gb=&brain= gives the setup's prompt, its starter name and its sheet (every step). The full checkout page is https://insidethemachine.ai/checkout (https://insidethemachine.ai/checkout?packs=&brain=&from=cart#start opens it with the cart's packs and brain). The folder: the folder is the sheet's starter folder plus MY-CART.md, one section per part (what it is, whether it fits, how to get it), at https://insidethemachine.ai/starter/cart/~.zip. Nothing to pay. Our app for every setup is Hermes Agent; what it already does is never sold here as an add-on.",
"Fits: the memory a part needs must be at most the usable share of your memory (75%). The cards mark it with three squares: runs well; tight (it fits, but the memory is nearly full); needs more. Text scores: Artificial Analysis Intelligence Index as a share of the best paid AI (Claude Opus 5.5), 2026-10-15. Sizes: Ollama library, 2026-10-26. Pictures and video: images.json, 2026-10-16. Voice and ears: speech.json, 2026-10-23. Downloads: Hugging Face, last 30 days. Unknown says unknown.",
"Can (text models): Writes code: it is made for code (models.json: the coding category; it is in the Code aisle too). Sees images: you can give it pictures or screenshots (models.json: the vision tag or category, or Image in its inputs). Hears audio: you can give it sound (models.json: the audio tag or category, or Audio in its inputs). Uses tools: it can call tools, so an app like Hermes Agent can let it search, read files and more (Ollama tag: tools, or the tool-calling category). Thinks: it can think step by step before it answers (Ollama tag: thinking). Very long docs: it takes in 1 million tokens or more at once (context_tokens.native, from its HF config.json). No badge means our data does not say it can; \"unknown: …\" means models.json has no capability tags (and for images, no inputs) for that model.",
"What Hermes Agent already does (no add-on needed): Read and write your files; Remember you; Search the web and read pages (its own docs disagree); Use your Gmail, calendar and Drive; Read your GitHub; Use your Notion; Use a web browser (its own docs disagree); Run commands; Look at pictures; Speak out loud; Do jobs on a schedule; Send out helper agents (its own docs disagree); Control your desktop (its own docs disagree); Skills (ready-made know-how) (its own docs disagree). (It loads skills itself; the Skills aisle sells skill packs for it to load.)"
],
"items": [],
"tables": []
},
{
"heading": "Text & chat (79)",
"level": 3,
"text": [
"Chats, writes and answers questions."
],
"items": [],
"tables": [
{
"columns": [
"Model",
"Maker",
"Needs",
"Quality",
"Business use",
"Released",
"Can",
"Good for"
],
"rows": [
{
"Model": "[GLM-5.3](https://huggingface.co/zai-org/GLM-5.3) (Mac only)",
"Maker": "Z.ai",
"Needs": "483.1 GB",
"Quality": "78% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-08-25",
"Can": "Writes code, Uses tools, Thinks, Very long docs; context 1,048,576 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[GLM-5.3-Flash](https://huggingface.co/zai-org/GLM-5.3-Flash) (Mac only)",
"Maker": "Z.ai",
"Needs": "233 GB",
"Quality": "73% of the best paid AI",
"Business use": "OK",
"Released": "2026-08-25",
"Can": "Writes code, Sees images, Uses tools, Thinks, Very long docs; context 1,048,576 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Qwen3.8-Flash-Next (125B-A6B)](https://huggingface.co/Qwen/Qwen3.8-Flash-Next)",
"Maker": "Alibaba Qwen",
"Needs": "140 GB",
"Quality": "69% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-08-24",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing, code and agent jobs."
},
{
"Model": "[Muse Spark Open 816B-A42B](https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B) (Mac only)",
"Maker": "Meta",
"Needs": "533 GB",
"Quality": "68% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-10-15",
"Can": "Writes code, Sees images, Hears audio, Uses tools, Thinks, Very long docs; context 1,000,000 tokens",
"Good for": "Good for: chat and writing, code, agent jobs and long documents."
},
{
"Model": "[Qwen3.8 27B](https://huggingface.co/Qwen/Qwen3.8-27B)",
"Maker": "Alibaba Qwen",
"Needs": "22.7 GB",
"Quality": "59% of the best paid AI",
"Business use": "OK",
"Released": "2026-08-05",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing, code and agent jobs."
},
{
"Model": "[MiniMax M3](https://huggingface.co/MiniMaxAI/MiniMax-M3) (Mac only)",
"Maker": "MiniMax",
"Needs": "302.5 GB",
"Quality": "51% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-06-02",
"Can": "Writes code, Sees images, Uses tools, Thinks, Very long docs; context 1,048,576 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Nemotron 3.5 Super VL 120B-A12B](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16)",
"Maker": "NVIDIA",
"Needs": "105.5 GB",
"Quality": "46% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-10-12",
"Can": "Writes code, Sees images, Uses tools, Thinks, Very long docs; context 1,048,576 tokens",
"Good for": "Good for: chat and writing, code, agent jobs, using tools, long documents and reading scans."
},
{
"Model": "[Kimi K2.7 Code](https://huggingface.co/moonshotai/Kimi-K2.7-Code) (Mac only)",
"Maker": "Moonshot AI",
"Needs": "739.2 GB",
"Quality": "45% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-06-11",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[K2 Horizon MoVA 36B-A4B](https://huggingface.co/IFM/K2-Horizon-MoVA-36B-A4B-GGUF) (Mac only)",
"Maker": "IFM (MBZUAI)",
"Needs": "28.1 GB",
"Quality": "44% of the best paid AI",
"Business use": "OK",
"Released": "2026-09-03",
"Can": "context unknown; unknown: images, tools, thinking",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Nemotron 3.5 Bolt 74B-A9B](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB)",
"Maker": "NVIDIA",
"Needs": "30.7 GB",
"Quality": "37% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-10-07",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing, code and agent jobs."
},
{
"Model": "[Qwen3.6 27B](https://huggingface.co/Qwen/Qwen3.6-27B)",
"Maker": "Alibaba Qwen",
"Needs": "21.5 GB",
"Quality": "37% of the best paid AI",
"Business use": "OK",
"Released": "2026-04-21",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing and code."
},
{
"Model": "[Qwen3.5 397B-A17B](https://huggingface.co/Qwen/Qwen3.5-397B-A17B) (Mac only)",
"Maker": "Alibaba Qwen",
"Needs": "282.7 GB",
"Quality": "37% of the best paid AI",
"Business use": "OK",
"Released": "2026-02-16",
"Can": "Sees images; context 262,144 tokens; unknown: tools, thinking",
"Good for": "Good for: chat and writing."
},
{
"Model": "[K2 Horizon 7B](https://huggingface.co/IFM/K2-Horizon-7B) (Mac only)",
"Maker": "IFM (MBZUAI)",
"Needs": "8.6 GB",
"Quality": "36% of the best paid AI",
"Business use": "OK",
"Released": "2026-09-01",
"Can": "context 524,288 tokens; unknown: images, tools, thinking",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Qwen3.6 35B-A3B](https://huggingface.co/Qwen/Qwen3.6-35B-A3B)",
"Maker": "Alibaba Qwen",
"Needs": "29.6 GB",
"Quality": "32% of the best paid AI",
"Business use": "OK",
"Released": "2026-04-15",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing and code."
},
{
"Model": "[Qwen3.5 122B-A10B](https://huggingface.co/Qwen/Qwen3.5-122B-A10B)",
"Maker": "Alibaba Qwen",
"Needs": "95.1 GB",
"Quality": "31% of the best paid AI",
"Business use": "OK",
"Released": "2026-02-24",
"Can": "Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Muse Glimmer 30B](https://huggingface.co/meta-models/Muse-Glimmer-30B)",
"Maker": "Meta",
"Needs": "22.7 GB",
"Quality": "30% of the best paid AI",
"Business use": "OK",
"Released": "2026-08-09",
"Can": "Sees images, Uses tools, Thinks; context 131,072 tokens",
"Good for": "Good for: chat and writing and agent jobs."
},
{
"Model": "[Gemma 4 26B-A4B](https://huggingface.co/google/gemma-4-26B-A4B-it)",
"Maker": "Google",
"Needs": "22.7 GB",
"Quality": "29% of the best paid AI",
"Business use": "OK",
"Released": "2026-03-11",
"Can": "Sees images, Hears audio, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Gemma 4 31B](https://huggingface.co/google/gemma-4-31B-it)",
"Maker": "Google",
"Needs": "25 GB",
"Quality": "26% of the best paid AI",
"Business use": "OK",
"Released": "2026-03-11",
"Can": "Sees images, Hears audio, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[GLM-4.7-Flash 30B-A3B](https://huggingface.co/zai-org/GLM-4.7-Flash)",
"Maker": "Z.ai",
"Needs": "23.8 GB",
"Quality": "26% of the best paid AI",
"Business use": "OK",
"Released": "2026-01-19",
"Can": "Writes code, Uses tools, Thinks; context 202,752 tokens",
"Good for": "Good for: chat and writing and code."
},
{
"Model": "[Gemma 4 12B](https://huggingface.co/google/gemma-4-12B-it)",
"Maker": "Google",
"Needs": "11.2 GB",
"Quality": "25% of the best paid AI",
"Business use": "OK",
"Released": "2026-05-23",
"Can": "Sees images, Hears audio, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Mistral Medium 3.5 128B](https://huggingface.co/mistralai/Mistral-Medium-3.5-128B)",
"Maker": "Mistral AI",
"Needs": "94 GB",
"Quality": "25% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-03-31",
"Can": "Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing and hard questions."
},
{
"Model": "[Qwen3.5 4B](https://huggingface.co/Qwen/Qwen3.5-4B)",
"Maker": "Alibaba Qwen",
"Needs": "5.8 GB",
"Quality": "23% of the best paid AI",
"Business use": "OK",
"Released": "2026-02-27",
"Can": "Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing. Small enough for a laptop or phone."
},
{
"Model": "[Qwen3.5 9B](https://huggingface.co/Qwen/Qwen3.5-9B)",
"Maker": "Alibaba Qwen",
"Needs": "9.6 GB",
"Quality": "23% of the best paid AI",
"Business use": "OK",
"Released": "2026-02-27",
"Can": "Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[MiniCPM5 2B](https://huggingface.co/openbmb/MiniCPM5-2B) (Mac only)",
"Maker": "OpenBMB",
"Needs": "3.9 GB",
"Quality": "22% of the best paid AI",
"Business use": "OK",
"Released": "2026-09-06",
"Can": "context 131,072 tokens; unknown: images, tools, thinking",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[Granite 4.2 30B](https://huggingface.co/ibm-granite/granite-4.2-30b)",
"Maker": "IBM",
"Needs": "22.7 GB",
"Quality": "22% of the best paid AI",
"Business use": "OK",
"Released": "2026-08-07",
"Can": "Uses tools; context 131,072 tokens; unknown: thinking",
"Good for": "Good for: chat and writing and using tools."
},
{
"Model": "[Nemotron 3.5 Lightning 30B-A3B](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16)",
"Maker": "NVIDIA",
"Needs": "30.7 GB",
"Quality": "22% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-08-01",
"Can": "Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing and agent jobs."
},
{
"Model": "Nemotron Cascade 2 30B-A3B",
"Maker": "NVIDIA",
"Needs": "29.6 GB",
"Quality": "20% of the best paid AI",
"Business use": "check the licence",
"Released": "unknown",
"Can": "Uses tools, Thinks; context unknown",
"Good for": "Good for: hard questions."
},
{
"Model": "[Mistral Small 4 119B](https://huggingface.co/mistralai/Mistral-Small-4-119B-2603) (Mac only)",
"Maker": "Mistral AI",
"Needs": "85 GB",
"Quality": "20% of the best paid AI",
"Business use": "OK",
"Released": "2026-01-23",
"Can": "Sees images, Very long docs; context 1,048,576 tokens; unknown: tools, thinking",
"Good for": "Good for: chat and writing."
},
{
"Model": "[gpt-oss-120b](https://huggingface.co/openai/gpt-oss-120b)",
"Maker": "OpenAI",
"Needs": "74.2 GB",
"Quality": "20% of the best paid AI",
"Business use": "OK",
"Released": "2025-08-04",
"Can": "Uses tools, Thinks; context 131,072 tokens",
"Good for": "Good for: chat and writing and hard questions."
},
{
"Model": "[Granite 4.2 8B](https://huggingface.co/ibm-granite/granite-4.2-8b)",
"Maker": "IBM",
"Needs": "8.1 GB",
"Quality": "19% of the best paid AI",
"Business use": "OK",
"Released": "2026-08-07",
"Can": "Uses tools; context 131,072 tokens; unknown: thinking",
"Good for": "Good for: chat and writing and using tools."
},
{
"Model": "[Nemotron 3 Nano Omni 30B-A3B](https://huggingface.co/nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-BF16)",
"Maker": "NVIDIA",
"Needs": "34.2 GB",
"Quality": "18% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-04-20",
"Can": "Sees images, Hears audio, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat."
},
{
"Model": "[North Mini Code 1.0 30B-A3B](https://huggingface.co/CohereLabs/North-Mini-Code-1.0)",
"Maker": "Cohere",
"Needs": "23.8 GB",
"Quality": "17% of the best paid AI",
"Business use": "OK",
"Released": "2026-06-05",
"Can": "Writes code, Uses tools, Thinks; context 500,000 tokens",
"Good for": "Good for: code."
},
{
"Model": "[gpt-oss-20b](https://huggingface.co/openai/gpt-oss-20b)",
"Maker": "OpenAI",
"Needs": "15.4 GB",
"Quality": "17% of the best paid AI",
"Business use": "OK",
"Released": "2025-08-04",
"Can": "Uses tools, Thinks; context 131,072 tokens",
"Good for": "Good for: chat and writing and hard questions."
},
{
"Model": "[Qwen3-Coder 30B-A3B](https://huggingface.co/Qwen/Qwen3-Coder-30B-A3B-Instruct)",
"Maker": "Alibaba Qwen",
"Needs": "23.8 GB",
"Quality": "17% of the best paid AI",
"Business use": "OK",
"Released": "2025-07-31",
"Can": "Writes code, Uses tools; context 262,144 tokens",
"Good for": "Good for: code."
},
{
"Model": "[Qwen3 30B-A3B (2507)](https://huggingface.co/Qwen/Qwen3-30B-A3B-Instruct-2507)",
"Maker": "Alibaba Qwen",
"Needs": "23.8 GB",
"Quality": "17% of the best paid AI",
"Business use": "OK",
"Released": "2025-07-28",
"Can": "Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Granite 4.2 3B](https://huggingface.co/ibm-granite/granite-4.2-3b)",
"Maker": "IBM",
"Needs": "4.5 GB",
"Quality": "16% of the best paid AI",
"Business use": "OK",
"Released": "2026-08-07",
"Can": "Uses tools; context 131,072 tokens; unknown: thinking",
"Good for": "Good for: using tools. Small enough for a laptop or phone."
},
{
"Model": "[Qwen3-Coder-Next (80B-A3B)](https://huggingface.co/Qwen/Qwen3-Coder-Next)",
"Maker": "Alibaba Qwen",
"Needs": "61.8 GB",
"Quality": "16% of the best paid AI",
"Business use": "OK",
"Released": "2026-01-30",
"Can": "Writes code, Uses tools; context 262,144 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Gemma 4 E4B](https://huggingface.co/google/gemma-4-E4B-it)",
"Maker": "Google",
"Needs": "9.6 GB",
"Quality": "15% of the best paid AI",
"Business use": "OK",
"Released": "2026-03-02",
"Can": "Sees images, Hears audio, Uses tools, Thinks; context 131,072 tokens",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[Devstral 2 123B](https://huggingface.co/mistralai/Devstral-2-123B-Instruct-2512)",
"Maker": "Mistral AI",
"Needs": "88.2 GB",
"Quality": "15% of the best paid AI",
"Business use": "check the licence",
"Released": "2025-11-28",
"Can": "Writes code, Uses tools; context 262,144 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Gemma 4 E2B](https://huggingface.co/google/gemma-4-E2B-it)",
"Maker": "Google",
"Needs": "7.3 GB",
"Quality": "14% of the best paid AI",
"Business use": "OK",
"Released": "2026-03-02",
"Can": "Sees images, Hears audio, Uses tools, Thinks; context 131,072 tokens",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[Qwen3-VL 8B](https://huggingface.co/Qwen/Qwen3-VL-8B-Instruct)",
"Maker": "Alibaba Qwen",
"Needs": "9 GB",
"Quality": "14% of the best paid AI",
"Business use": "OK",
"Released": "2025-10-11",
"Can": "Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat."
},
{
"Model": "[Hermes 4 70B (Llama 3.1)](https://huggingface.co/NousResearch/Hermes-4-70B) (Mac only)",
"Maker": "Nous Research",
"Needs": "50.9 GB",
"Quality": "14% of the best paid AI",
"Business use": "check the licence",
"Released": "2025-08-18",
"Can": "context 131,072 tokens; unknown: images, tools, thinking",
"Good for": "Good for: chat and writing."
},
{
"Model": "[DeepSeek-R1-0528-Qwen3-8B](https://huggingface.co/deepseek-ai/DeepSeek-R1-0528-Qwen3-8B)",
"Maker": "DeepSeek",
"Needs": "8 GB",
"Quality": "14% of the best paid AI",
"Business use": "OK",
"Released": "2025-05-29",
"Can": "Uses tools, Thinks; context 131,072 tokens",
"Good for": "Good for: hard questions."
},
{
"Model": "[LFM2.5 8B-A1B](https://huggingface.co/LiquidAI/LFM2.5-8B-A1B)",
"Maker": "Liquid AI",
"Needs": "8 GB",
"Quality": "13% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-05-28",
"Can": "Uses tools, Thinks; context 128,000 tokens",
"Good for": "Good for: using tools. Small enough for a laptop or phone."
},
{
"Model": "[Devstral Small 2 24B](https://huggingface.co/mistralai/Devstral-Small-2-24B-Instruct-2512)",
"Maker": "Mistral AI",
"Needs": "19.2 GB",
"Quality": "13% of the best paid AI",
"Business use": "OK",
"Released": "2025-11-28",
"Can": "Writes code, Sees images, Uses tools; context 393,216 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Llama 3.3 70B](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct)",
"Maker": "Meta",
"Needs": "51.4 GB",
"Quality": "13% of the best paid AI",
"Business use": "check the licence",
"Released": "2024-11-26",
"Can": "Uses tools; context unknown",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Qwen3.5 2B](https://huggingface.co/Qwen/Qwen3.5-2B)",
"Maker": "Alibaba Qwen",
"Needs": "4.2 GB",
"Quality": "12% of the best paid AI",
"Business use": "OK",
"Released": "2026-02-28",
"Can": "Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[OLMo 3.1 32B Think](https://huggingface.co/allenai/Olmo-3.1-32B-Think)",
"Maker": "Ai2",
"Needs": "23.8 GB",
"Quality": "12% of the best paid AI",
"Business use": "OK",
"Released": "2025-12-10",
"Can": "Uses tools, Thinks; context 65,536 tokens",
"Good for": "Good for: hard questions."
},
{
"Model": "[Mistral Small 3.2 24B](https://huggingface.co/mistralai/Mistral-Small-3.2-24B-Instruct-2506)",
"Maker": "Mistral AI",
"Needs": "19.2 GB",
"Quality": "12% of the best paid AI",
"Business use": "OK",
"Released": "2025-06-19",
"Can": "Sees images, Uses tools; context 131,072 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Llama 3.1 8B](https://huggingface.co/meta-llama/Llama-3.1-8B-Instruct)",
"Maker": "Meta",
"Needs": "7.6 GB",
"Quality": "12% of the best paid AI",
"Business use": "check the licence",
"Released": "2024-07-18",
"Can": "Uses tools; context unknown",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Qwen3.5 0.8B](https://huggingface.co/Qwen/Qwen3.5-0.8B)",
"Maker": "Alibaba Qwen",
"Needs": "2.6 GB",
"Quality": "11% of the best paid AI",
"Business use": "OK",
"Released": "2026-02-28",
"Can": "Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[OLMo 3.1 32B Instruct](https://huggingface.co/allenai/Olmo-3.1-32B-Instruct)",
"Maker": "Ai2",
"Needs": "23.8 GB",
"Quality": "11% of the best paid AI",
"Business use": "OK",
"Released": "2025-12-10",
"Can": "Uses tools, Thinks; context 65,536 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[MiniCPM-V 4.6 (1B)](https://huggingface.co/openbmb/MiniCPM-V-4.6)",
"Maker": "OpenBMB",
"Needs": "3.8 GB",
"Quality": "10% of the best paid AI",
"Business use": "OK",
"Released": "2026-04-13",
"Can": "Sees images; context 262,144 tokens",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[OLMo 3 7B Instruct](https://huggingface.co/allenai/Olmo-3-7B-Instruct)",
"Maker": "Ai2",
"Needs": "7.2 GB",
"Quality": "10% of the best paid AI",
"Business use": "OK",
"Released": "2025-11-19",
"Can": "Uses tools, Thinks; context 65,536 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Ministral 3 8B](https://huggingface.co/mistralai/Ministral-3-8B-Instruct-2512)",
"Maker": "Mistral AI",
"Needs": "8.9 GB",
"Quality": "10% of the best paid AI",
"Business use": "OK",
"Released": "2025-10-31",
"Can": "Sees images, Uses tools; context 262,144 tokens",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[Ministral 3 14B](https://huggingface.co/mistralai/Ministral-3-14B-Instruct-2512)",
"Maker": "Mistral AI",
"Needs": "12.5 GB",
"Quality": "10% of the best paid AI",
"Business use": "OK",
"Released": "2025-10-31",
"Can": "Sees images, Uses tools; context 262,144 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Qwen3 8B](https://huggingface.co/Qwen/Qwen3-8B)",
"Maker": "Alibaba Qwen",
"Needs": "8 GB",
"Quality": "10% of the best paid AI",
"Business use": "OK",
"Released": "2025-04-27",
"Can": "Uses tools, Thinks; context 40,960 tokens",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Phi-4 14B](https://huggingface.co/microsoft/phi-4)",
"Maker": "Microsoft",
"Needs": "12.5 GB",
"Quality": "10% of the best paid AI",
"Business use": "OK",
"Released": "2024-12-11",
"Can": "context 16,384 tokens; unknown: tools, thinking",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Llama 3.2 3B](https://huggingface.co/meta-llama/Llama-3.2-3B-Instruct)",
"Maker": "Meta",
"Needs": "4.3 GB",
"Quality": "10% of the best paid AI",
"Business use": "check the licence",
"Released": "2024-09-18",
"Can": "Uses tools; context unknown",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[LFM2.5 1.2B Thinking](https://huggingface.co/LiquidAI/LFM2.5-1.2B-Thinking)",
"Maker": "Liquid AI",
"Needs": "2.8 GB",
"Quality": "9% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-01-20",
"Can": "Uses tools, Thinks; context 128,000 tokens",
"Good for": "Good for: hard questions. Small enough for a laptop or phone."
},
{
"Model": "[Gemma 3 27B](https://huggingface.co/google/gemma-3-27b-it)",
"Maker": "Google",
"Needs": "21.5 GB",
"Quality": "9% of the best paid AI",
"Business use": "check the licence",
"Released": "2025-03-01",
"Can": "Sees images; context unknown",
"Good for": "Good for: chat and writing."
},
{
"Model": "[Nemotron 3 Nano 4B](https://huggingface.co/nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16)",
"Maker": "NVIDIA",
"Needs": "5.2 GB",
"Quality": "8% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-03-07",
"Can": "Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[Ministral 3 3B](https://huggingface.co/mistralai/Ministral-3-3B-Instruct-2512)",
"Maker": "Mistral AI",
"Needs": "5.4 GB",
"Quality": "8% of the best paid AI",
"Business use": "OK",
"Released": "2025-10-31",
"Can": "Sees images, Uses tools; context 262,144 tokens",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[Gemma 3 4B](https://huggingface.co/google/gemma-3-4b-it)",
"Maker": "Google",
"Needs": "5.9 GB",
"Quality": "8% of the best paid AI",
"Business use": "check the licence",
"Released": "2025-02-20",
"Can": "Sees images; context unknown",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[Nanbeige4.3 33B](https://huggingface.co/Nanbeige/Nanbeige4.3-33B) (Mac only)",
"Maker": "Nanbeige",
"Needs": "25.9 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-10-24",
"Can": "Writes code, Uses tools, Thinks; context 524,288 tokens",
"Good for": "Good for: code, hard questions, agent jobs, using tools and long documents."
},
{
"Model": "[Recursive Flint 40B-A2.5B](https://huggingface.co/RecursiveSI/flint)",
"Maker": "Recursive",
"Needs": "28 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-10-21",
"Can": "Writes code, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: code, hard questions, agent jobs and using tools."
},
{
"Model": "[LFM2.5 24B-A2B](https://huggingface.co/LiquidAI/LFM2.5-24B-A2B)",
"Maker": "Liquid AI",
"Needs": "19.2 GB",
"Quality": "unknown",
"Business use": "check the licence",
"Released": "2026-10-12",
"Can": "Uses tools, Thinks; context 131,072 tokens",
"Good for": "Good for: chat and writing, hard questions, agent jobs and using tools."
},
{
"Model": "[Nemotron 3.5 Bolt E4B](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-E4B-BF16)",
"Maker": "NVIDIA",
"Needs": "11.1 GB",
"Quality": "unknown",
"Business use": "check the licence",
"Released": "2026-10-05",
"Can": "Sees images, Uses tools; context 65,536 tokens",
"Good for": "Good for: using tools. Small enough for a laptop or phone."
},
{
"Model": "[Ornith 1.5 9B](https://huggingface.co/ornith-ai/Ornith-1.5-9B-GGUF)",
"Maker": "Ornith AI",
"Needs": "8.8 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-08-19",
"Can": "Writes code, Sees images; context unknown",
"Good for": "Good for: code."
},
{
"Model": "[Ornith 1.5 35B-A3B](https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B-GGUF)",
"Maker": "Ornith AI",
"Needs": "27.1 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-08-18",
"Can": "Writes code, Sees images; context unknown",
"Good for": "Good for: code."
},
{
"Model": "[Apertus 1.5 8B](https://huggingface.co/swiss-ai/Apertus-v1.5-8B) (Mac only)",
"Maker": "Swiss AI Initiative",
"Needs": "7.8 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-07-24",
"Can": "Sees images, Hears audio, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing and long documents."
},
{
"Model": "[Laguna S 2.1 118B-A8B](https://huggingface.co/poolside/Laguna-S-2.1)",
"Maker": "Poolside",
"Needs": "112.4 GB",
"Quality": "unknown",
"Business use": "check the licence",
"Released": "2026-07-13",
"Can": "Writes code, Uses tools, Thinks, Very long docs; context 1,048,576 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Laguna XS 2.1 33B-A3B](https://huggingface.co/poolside/Laguna-XS-2.1)",
"Maker": "Poolside",
"Needs": "25 GB",
"Quality": "unknown",
"Business use": "check the licence",
"Released": "2026-06-20",
"Can": "Writes code, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[GLM-OCR](https://huggingface.co/zai-org/GLM-OCR)",
"Maker": "Z.ai",
"Needs": "5.1 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-01-30",
"Can": "Sees images, Uses tools; context 131,072 tokens",
"Good for": "Good for: reading scans."
},
{
"Model": "[DeepSeek-OCR 3B](https://huggingface.co/deepseek-ai/DeepSeek-OCR)",
"Maker": "DeepSeek",
"Needs": "4.8 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2025-10-17",
"Can": "Sees images; context 8,192 tokens",
"Good for": "Good for: reading scans."
},
{
"Model": "[FunctionGemma 270M](https://huggingface.co/google/functiongemma-270m-it)",
"Maker": "Google",
"Needs": "2.2 GB",
"Quality": "unknown",
"Business use": "check the licence",
"Released": "2025-10-08",
"Can": "Uses tools; context unknown",
"Good for": "Good for: using tools. Small enough for a laptop or phone."
},
{
"Model": "[Hermes 4 14B](https://huggingface.co/NousResearch/Hermes-4-14B) (Mac only)",
"Maker": "Nous Research",
"Needs": "12.3 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2025-08-30",
"Can": "context 40,960 tokens; unknown: images, tools, thinking",
"Good for": "Good for: chat and writing."
},
{
"Model": "[SmolLM3 3B](https://huggingface.co/HuggingFaceTB/SmolLM3-3B) (Mac only)",
"Maker": "Hugging Face",
"Needs": "4.2 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2025-07-08",
"Can": "context 65,536 tokens; unknown: images, tools, thinking",
"Good for": "Good for: chat. Small enough for a laptop or phone."
},
{
"Model": "[Hermes 3 8B (Llama 3.1)](https://huggingface.co/NousResearch/Hermes-3-Llama-3.1-8B)",
"Maker": "Nous Research",
"Needs": "7.6 GB",
"Quality": "unknown",
"Business use": "check the licence",
"Released": "2024-07-28",
"Can": "Uses tools; context 131,072 tokens",
"Good for": "Good for: chat and writing."
}
]
}
]
},
{
"heading": "Code (23)",
"level": 3,
"text": [
"Writes and fixes code."
],
"items": [],
"tables": [
{
"columns": [
"Model",
"Maker",
"Needs",
"Quality",
"Business use",
"Released",
"Can",
"Good for"
],
"rows": [
{
"Model": "[GLM-5.3](https://huggingface.co/zai-org/GLM-5.3) (Mac only)",
"Maker": "Z.ai",
"Needs": "483.1 GB",
"Quality": "78% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-08-25",
"Can": "Writes code, Uses tools, Thinks, Very long docs; context 1,048,576 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[GLM-5.3-Flash](https://huggingface.co/zai-org/GLM-5.3-Flash) (Mac only)",
"Maker": "Z.ai",
"Needs": "233 GB",
"Quality": "73% of the best paid AI",
"Business use": "OK",
"Released": "2026-08-25",
"Can": "Writes code, Sees images, Uses tools, Thinks, Very long docs; context 1,048,576 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Qwen3.8-Flash-Next (125B-A6B)](https://huggingface.co/Qwen/Qwen3.8-Flash-Next)",
"Maker": "Alibaba Qwen",
"Needs": "140 GB",
"Quality": "69% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-08-24",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing, code and agent jobs."
},
{
"Model": "[Muse Spark Open 816B-A42B](https://huggingface.co/meta-models/Muse-Spark-Open-816B-A42B) (Mac only)",
"Maker": "Meta",
"Needs": "533 GB",
"Quality": "68% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-10-15",
"Can": "Writes code, Sees images, Hears audio, Uses tools, Thinks, Very long docs; context 1,000,000 tokens",
"Good for": "Good for: chat and writing, code, agent jobs and long documents."
},
{
"Model": "[Qwen3.8 27B](https://huggingface.co/Qwen/Qwen3.8-27B)",
"Maker": "Alibaba Qwen",
"Needs": "22.7 GB",
"Quality": "59% of the best paid AI",
"Business use": "OK",
"Released": "2026-08-05",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing, code and agent jobs."
},
{
"Model": "[MiniMax M3](https://huggingface.co/MiniMaxAI/MiniMax-M3) (Mac only)",
"Maker": "MiniMax",
"Needs": "302.5 GB",
"Quality": "51% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-06-02",
"Can": "Writes code, Sees images, Uses tools, Thinks, Very long docs; context 1,048,576 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Nemotron 3.5 Super VL 120B-A12B](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Super-VL-120B-A12B-BF16)",
"Maker": "NVIDIA",
"Needs": "105.5 GB",
"Quality": "46% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-10-12",
"Can": "Writes code, Sees images, Uses tools, Thinks, Very long docs; context 1,048,576 tokens",
"Good for": "Good for: chat and writing, code, agent jobs, using tools, long documents and reading scans."
},
{
"Model": "[Kimi K2.7 Code](https://huggingface.co/moonshotai/Kimi-K2.7-Code) (Mac only)",
"Maker": "Moonshot AI",
"Needs": "739.2 GB",
"Quality": "45% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-06-11",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Nemotron 3.5 Bolt 74B-A9B](https://huggingface.co/nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-24GB)",
"Maker": "NVIDIA",
"Needs": "30.7 GB",
"Quality": "37% of the best paid AI",
"Business use": "check the licence",
"Released": "2026-10-07",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing, code and agent jobs."
},
{
"Model": "[Qwen3.6 27B](https://huggingface.co/Qwen/Qwen3.6-27B)",
"Maker": "Alibaba Qwen",
"Needs": "21.5 GB",
"Quality": "37% of the best paid AI",
"Business use": "OK",
"Released": "2026-04-21",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing and code."
},
{
"Model": "[Qwen3.6 35B-A3B](https://huggingface.co/Qwen/Qwen3.6-35B-A3B)",
"Maker": "Alibaba Qwen",
"Needs": "29.6 GB",
"Quality": "32% of the best paid AI",
"Business use": "OK",
"Released": "2026-04-15",
"Can": "Writes code, Sees images, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: chat and writing and code."
},
{
"Model": "[GLM-4.7-Flash 30B-A3B](https://huggingface.co/zai-org/GLM-4.7-Flash)",
"Maker": "Z.ai",
"Needs": "23.8 GB",
"Quality": "26% of the best paid AI",
"Business use": "OK",
"Released": "2026-01-19",
"Can": "Writes code, Uses tools, Thinks; context 202,752 tokens",
"Good for": "Good for: chat and writing and code."
},
{
"Model": "[North Mini Code 1.0 30B-A3B](https://huggingface.co/CohereLabs/North-Mini-Code-1.0)",
"Maker": "Cohere",
"Needs": "23.8 GB",
"Quality": "17% of the best paid AI",
"Business use": "OK",
"Released": "2026-06-05",
"Can": "Writes code, Uses tools, Thinks; context 500,000 tokens",
"Good for": "Good for: code."
},
{
"Model": "[Qwen3-Coder 30B-A3B](https://huggingface.co/Qwen/Qwen3-Coder-30B-A3B-Instruct)",
"Maker": "Alibaba Qwen",
"Needs": "23.8 GB",
"Quality": "17% of the best paid AI",
"Business use": "OK",
"Released": "2025-07-31",
"Can": "Writes code, Uses tools; context 262,144 tokens",
"Good for": "Good for: code."
},
{
"Model": "[Qwen3-Coder-Next (80B-A3B)](https://huggingface.co/Qwen/Qwen3-Coder-Next)",
"Maker": "Alibaba Qwen",
"Needs": "61.8 GB",
"Quality": "16% of the best paid AI",
"Business use": "OK",
"Released": "2026-01-30",
"Can": "Writes code, Uses tools; context 262,144 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Devstral 2 123B](https://huggingface.co/mistralai/Devstral-2-123B-Instruct-2512)",
"Maker": "Mistral AI",
"Needs": "88.2 GB",
"Quality": "15% of the best paid AI",
"Business use": "check the licence",
"Released": "2025-11-28",
"Can": "Writes code, Uses tools; context 262,144 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Devstral Small 2 24B](https://huggingface.co/mistralai/Devstral-Small-2-24B-Instruct-2512)",
"Maker": "Mistral AI",
"Needs": "19.2 GB",
"Quality": "13% of the best paid AI",
"Business use": "OK",
"Released": "2025-11-28",
"Can": "Writes code, Sees images, Uses tools; context 393,216 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Nanbeige4.3 33B](https://huggingface.co/Nanbeige/Nanbeige4.3-33B) (Mac only)",
"Maker": "Nanbeige",
"Needs": "25.9 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-10-24",
"Can": "Writes code, Uses tools, Thinks; context 524,288 tokens",
"Good for": "Good for: code, hard questions, agent jobs, using tools and long documents."
},
{
"Model": "[Recursive Flint 40B-A2.5B](https://huggingface.co/RecursiveSI/flint)",
"Maker": "Recursive",
"Needs": "28 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-10-21",
"Can": "Writes code, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: code, hard questions, agent jobs and using tools."
},
{
"Model": "[Ornith 1.5 9B](https://huggingface.co/ornith-ai/Ornith-1.5-9B-GGUF)",
"Maker": "Ornith AI",
"Needs": "8.8 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-08-19",
"Can": "Writes code, Sees images; context unknown",
"Good for": "Good for: code."
},
{
"Model": "[Ornith 1.5 35B-A3B](https://huggingface.co/ornith-ai/Ornith-1.5-35B-A3B-GGUF)",
"Maker": "Ornith AI",
"Needs": "27.1 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-08-18",
"Can": "Writes code, Sees images; context unknown",
"Good for": "Good for: code."
},
{
"Model": "[Laguna S 2.1 118B-A8B](https://huggingface.co/poolside/Laguna-S-2.1)",
"Maker": "Poolside",
"Needs": "112.4 GB",
"Quality": "unknown",
"Business use": "check the licence",
"Released": "2026-07-13",
"Can": "Writes code, Uses tools, Thinks, Very long docs; context 1,048,576 tokens",
"Good for": "Good for: code and agent jobs."
},
{
"Model": "[Laguna XS 2.1 33B-A3B](https://huggingface.co/poolside/Laguna-XS-2.1)",
"Maker": "Poolside",
"Needs": "25 GB",
"Quality": "unknown",
"Business use": "check the licence",
"Released": "2026-06-20",
"Can": "Writes code, Uses tools, Thinks; context 262,144 tokens",
"Good for": "Good for: code and agent jobs."
}
]
}
]
},
{
"heading": "Images (24)",
"level": 3,
"text": [
"Makes pictures from your words."
],
"items": [],
"tables": [
{
"columns": [
"Model",
"Maker",
"Needs",
"Quality",
"Business use",
"Released",
"Good for"
],
"rows": [
{
"Model": "[Qwen-Image 2.1](https://huggingface.co/Qwen/Qwen-Image-2.1)",
"Maker": "Alibaba (Qwen)",
"Needs": "46 GB",
"Quality": "#1 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "No",
"Released": "2026-09-20",
"Good for": "Good for: Best open quality; text, edits, transparency."
},
{
"Model": "[Qwen-Image 2512](https://huggingface.co/Qwen/Qwen-Image-2512)",
"Maker": "Alibaba (Qwen)",
"Needs": "unknown",
"Quality": "#1 of 2 (LMArena Text-to-Image Arena)",
"Business use": "OK",
"Released": "2025-12-30",
"Good for": "Good for: Realistic people, commercial-friendly license."
},
{
"Model": "[Qwen-Image-Edit 2511](https://huggingface.co/Qwen/Qwen-Image-Edit-2511)",
"Maker": "Alibaba (Qwen)",
"Needs": "unknown",
"Quality": "#1 of 3 (Artificial Analysis Image Editing Arena (AA-Image-Editing v2.1))",
"Business use": "OK",
"Released": "2025-12-17",
"Good for": "Good for: Text-instructed edits keeping people consistent."
},
{
"Model": "[Krea 2 Turbo](https://huggingface.co/krea/Krea-2-Turbo)",
"Maker": "Krea",
"Needs": "unknown",
"Quality": "#2 of 2 (LMArena Text-to-Image Arena)",
"Business use": "OK",
"Released": "2026-06-22",
"Good for": "Good for: Wide range of creative styles."
},
{
"Model": "[Ideogram 4](https://huggingface.co/ideogram-ai/ideogram-4-fp8)",
"Maker": "Ideogram",
"Needs": "unknown",
"Quality": "#2 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "No",
"Released": "2026-06-03",
"Good for": "Good for: Typography, posters and design layouts."
},
{
"Model": "[FIBO Edit](https://huggingface.co/briaai/Fibo-Edit)",
"Maker": "Bria",
"Needs": "unknown",
"Quality": "#2 of 3 (Artificial Analysis Image Editing Arena (AA-Image-Editing v2.1))",
"Business use": "No",
"Released": "2026-01-20",
"Good for": "Good for: Controlled edits from text instructions."
},
{
"Model": "[FLUX.2 [dev]](https://huggingface.co/black-forest-labs/FLUX.2-dev)",
"Maker": "Black Forest Labs",
"Needs": "24 GB",
"Quality": "#3 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "No",
"Released": "2025-11-25",
"Good for": "Good for: High-end photoreal images and multi-reference edits."
},
{
"Model": "[FLUX.1 Kontext [dev]](https://huggingface.co/black-forest-labs/FLUX.1-Kontext-dev)",
"Maker": "Black Forest Labs",
"Needs": "6 GB",
"Quality": "#3 of 3 (Artificial Analysis Image Editing Arena (AA-Image-Editing v2.1))",
"Business use": "No",
"Released": "2025-06-26",
"Good for": "Good for: Simple text-instructed edits, wide tool support."
},
{
"Model": "[Ming-Image 0.1 Design](https://huggingface.co/inclusionAI/Ming-Image-0.1-Design)",
"Maker": "inclusionAI (Ant Group)",
"Needs": "24 GB",
"Quality": "#4 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "OK",
"Released": "2026-09-22",
"Good for": "Good for: Posters, cards and UI with text."
},
{
"Model": "[HiDream-O1-Image](https://huggingface.co/HiDream-ai/HiDream-O1-Image)",
"Maker": "HiDream.ai",
"Needs": "unknown",
"Quality": "#5 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "OK",
"Released": "2026-05-08",
"Good for": "Good for: Strong prompt following, MIT licensed."
},
{
"Model": "[Z-Image Turbo](https://huggingface.co/Tongyi-MAI/Z-Image-Turbo)",
"Maker": "Alibaba (Tongyi-MAI)",
"Needs": "4 GB",
"Quality": "#6 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "OK",
"Released": "2025-12-02",
"Good for": "Good for: Fast photoreal images on modest hardware."
},
{
"Model": "[FLUX.2 [klein] 9B](https://huggingface.co/black-forest-labs/FLUX.2-klein-9B)",
"Maker": "Black Forest Labs",
"Needs": "29 GB",
"Quality": "#7 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "No",
"Released": "2026-01-15",
"Good for": "Good for: Fast generation and editing in one model."
},
{
"Model": "[ERNIE-Image Turbo](https://huggingface.co/baidu/ERNIE-Image-Turbo)",
"Maker": "Baidu",
"Needs": "4 GB",
"Quality": "#8 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "OK",
"Released": "2026-04-15",
"Good for": "Good for: Vivid, high-contrast images, Apache licensed."
},
{
"Model": "[Z-Image (Base)](https://huggingface.co/Tongyi-MAI/Z-Image)",
"Maker": "Alibaba (Tongyi-MAI)",
"Needs": "4 GB",
"Quality": "#9 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "OK",
"Released": "2026-01-27",
"Good for": "Good for: Fine-tuning and LoRA base for Z-Image."
},
{
"Model": "[GLM-Image](https://huggingface.co/zai-org/GLM-Image)",
"Maker": "Z.ai",
"Needs": "23 GB",
"Quality": "#10 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "OK",
"Released": "2026-01-13",
"Good for": "Good for: Accurate text rendering inside images."
},
{
"Model": "[LongCat-Image](https://huggingface.co/meituan-longcat/LongCat-Image)",
"Maker": "Meituan",
"Needs": "17 GB",
"Quality": "#11 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "OK",
"Released": "2025-12-05",
"Good for": "Good for: Small bilingual model, Apache licensed."
},
{
"Model": "[FIBO](https://huggingface.co/briaai/FIBO)",
"Maker": "Bria",
"Needs": "unknown",
"Quality": "#12 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "No",
"Released": "2025-10-30",
"Good for": "Good for: Precise control through structured JSON prompts."
},
{
"Model": "[FLUX.2 [klein] 4B](https://huggingface.co/black-forest-labs/FLUX.2-klein-4B)",
"Maker": "Black Forest Labs",
"Needs": "13 GB",
"Quality": "#13 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "OK",
"Released": "2026-01-15",
"Good for": "Good for: Small, fast, commercial-friendly generation and edits."
},
{
"Model": "[FLUX.1 Krea [dev]](https://huggingface.co/black-forest-labs/FLUX.1-Krea-dev)",
"Maker": "Black Forest Labs and Krea",
"Needs": "unknown",
"Quality": "#14 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "No",
"Released": "2025-07-31",
"Good for": "Good for: Photorealism with a less 'AI' look."
},
{
"Model": "[FLUX.1 [dev]](https://huggingface.co/black-forest-labs/FLUX.1-dev)",
"Maker": "Black Forest Labs",
"Needs": "6 GB",
"Quality": "#15 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "No",
"Released": "2024-08-01",
"Good for": "Good for: Legacy workhorse with very wide tool support."
},
{
"Model": "[FLUX.1 [schnell]](https://huggingface.co/black-forest-labs/FLUX.1-schnell)",
"Maker": "Black Forest Labs",
"Needs": "6 GB",
"Quality": "#16 of 16 (Artificial Analysis Text to Image Arena (AA-Image-T2I v2.1))",
"Business use": "OK",
"Released": "2024-08-02",
"Good for": "Good for: Fast, Apache-licensed legacy FLUX."
},
{
"Model": "[FLUX.3 Image [klein]](https://huggingface.co/black-forest-labs/flux-3-image-klein)",
"Maker": "Black Forest Labs",
"Needs": "25 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-10-13",
"Good for": "Good for: Small FLUX for making and editing pictures, up to 10 reference images."
},
{
"Model": "[Qwen-Image 2.1 Turbo](https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo)",
"Maker": "Alibaba (Qwen)",
"Needs": "unknown",
"Quality": "unknown",
"Business use": "No",
"Released": "2026-10-09",
"Good for": "Good for: Faster 8-step Qwen-Image 2.1."
},
{
"Model": "[Stable Diffusion XL 1.0](https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0)",
"Maker": "Stability AI",
"Needs": "8 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2023-07-25",
"Good for": "Good for: Low-memory machines and huge community model library."
}
]
}
]
},
{
"heading": "Video (6)",
"level": 3,
"text": [
"Makes short clips."
],
"items": [],
"tables": [
{
"columns": [
"Model",
"Maker",
"Needs",
"Quality",
"Business use",
"Released",
"Good for"
],
"rows": [
{
"Model": "[MiniMax H3](https://huggingface.co/MiniMaxAI/MiniMax-H3)",
"Maker": "MiniMax",
"Needs": "42.5 GB",
"Quality": "#1 of 6 (Artificial Analysis Image to Video Arena, silent (AA-Video-I2V-Silent v1.0))",
"Business use": "OK",
"Released": "2026-08-03",
"Good for": "Good for: Best open video, with native stereo audio."
},
{
"Model": "[LTX-2.5](https://huggingface.co/Lightricks/LTX-2.5)",
"Maker": "Lightricks",
"Needs": "unknown",
"Quality": "#2 of 6 (Artificial Analysis Image to Video Arena, silent (AA-Video-I2V-Silent v1.0))",
"Business use": "OK",
"Released": "2026-08-11",
"Good for": "Good for: Video with audio, fast distilled mode."
},
{
"Model": "[LTX-2.3](https://huggingface.co/Lightricks/LTX-2.3)",
"Maker": "Lightricks",
"Needs": "unknown",
"Quality": "#3 of 6 (Artificial Analysis Image to Video Arena, silent (AA-Video-I2V-Silent v1.0))",
"Business use": "OK",
"Released": "2026-03-05",
"Good for": "Good for: Video with audio, built into Draw Things."
},
{
"Model": "[HunyuanVideo 1.5](https://huggingface.co/tencent/HunyuanVideo-1.5)",
"Maker": "Tencent",
"Needs": "14 GB",
"Quality": "#4 of 6 (Artificial Analysis Image to Video Arena, silent (AA-Video-I2V-Silent v1.0))",
"Business use": "OK",
"Released": "2025-11-20",
"Good for": "Good for: Good video on consumer GPUs."
},
{
"Model": "[Wan 2.2 A14B](https://huggingface.co/Wan-AI/Wan2.2-T2V-A14B)",
"Maker": "Alibaba (Wan)",
"Needs": "80 GB",
"Quality": "#5 of 6 (Artificial Analysis Image to Video Arena, silent (AA-Video-I2V-Silent v1.0))",
"Business use": "OK",
"Released": "2025-07-28",
"Good for": "Good for: High-quality 5 s clips, Apache licensed."
},
{
"Model": "[Wan 2.2 TI2V 5B](https://huggingface.co/Wan-AI/Wan2.2-TI2V-5B)",
"Maker": "Alibaba (Wan)",
"Needs": "24 GB",
"Quality": "#6 of 6 (Artificial Analysis Image to Video Arena, silent (AA-Video-I2V-Silent v1.0))",
"Business use": "OK",
"Released": "2025-07-28",
"Good for": "Good for: 720p video on one consumer GPU."
}
]
}
]
},
{
"heading": "Voice (22)",
"level": 3,
"text": [
"Speech out: reads text out loud."
],
"items": [],
"tables": [
{
"columns": [
"Model",
"Maker",
"Needs",
"Quality",
"Business use",
"Released",
"Good for"
],
"rows": [
{
"Model": "[Fish Audio Drama 3 (preview)](https://huggingface.co/fishaudio/drama-3)",
"Maker": "Fish Audio",
"Needs": "28 GB",
"Quality": "#1 of 12 (Artificial Analysis Speech Arena)",
"Business use": "check the licence",
"Released": "2026-10-01",
"Good for": "Good for: Best open voice quality, if you have RAM."
},
{
"Model": "[Breeze TTS 2](https://huggingface.co/BreezeBlue/Breeze-TTS-2)",
"Maker": "BreezeBlue",
"Needs": "8 GB",
"Quality": "#2 of 12 (Artificial Analysis Speech Arena)",
"Business use": "No",
"Released": "2026-08-25",
"Good for": "Good for: Top open voice cloning for personal projects."
},
{
"Model": "[Fish Audio S2 Pro](https://huggingface.co/fishaudio/s2-pro)",
"Maker": "Fish Audio",
"Needs": "8 GB",
"Quality": "#3 of 12 (Artificial Analysis Speech Arena)",
"Business use": "No",
"Released": "2026-03-09",
"Good for": "Good for: Expressive multilingual speech with inline emotion tags."
},
{
"Model": "[Step-Audio-EditX](https://huggingface.co/stepfun-ai/Step-Audio-EditX)",
"Maker": "StepFun",
"Needs": "12 GB",
"Quality": "#4 of 12 (Artificial Analysis Speech Arena)",
"Business use": "OK",
"Released": "2025-11-12",
"Good for": "Good for: Editing emotion and style of speech."
},
{
"Model": "[Voxtral TTS (4B)](https://huggingface.co/mistralai/Voxtral-4B-TTS-2603)",
"Maker": "Mistral AI",
"Needs": "16 GB",
"Quality": "#5 of 12 (Artificial Analysis Speech Arena)",
"Business use": "No",
"Released": "2026-03-26",
"Good for": "Good for: Low-latency voice agents in European languages."
},
{
"Model": "[Magpie TTS Multilingual 357M](https://huggingface.co/nvidia/magpie_tts_multilingual_357m)",
"Maker": "NVIDIA",
"Needs": "2 GB",
"Quality": "#6 of 12 (Artificial Analysis Speech Arena)",
"Business use": "OK",
"Released": "unknown",
"Good for": "Good for: Commercial multilingual voices for voice agents."
},
{
"Model": "[Kokoro 82M v1.0](https://huggingface.co/hexgrad/Kokoro-82M)",
"Maker": "hexgrad",
"Needs": "0.5 GB",
"Quality": "#7 of 12 (Artificial Analysis Speech Arena)",
"Business use": "OK",
"Released": "2025-01-27",
"Good for": "Good for: Fast, tiny, free for commercial use."
},
{
"Model": "[Maya1](https://huggingface.co/maya-research/maya1)",
"Maker": "Maya Research",
"Needs": "8 GB",
"Quality": "#8 of 12 (Artificial Analysis Speech Arena)",
"Business use": "OK",
"Released": "2025-10-18",
"Good for": "Good for: Design a voice by describing it."
},
{
"Model": "[Higgs TTS 3 (4B)](https://huggingface.co/bosonai/higgs-tts-3-4b)",
"Maker": "Boson AI",
"Needs": "11 GB",
"Quality": "#9 of 12 (Artificial Analysis Speech Arena)",
"Business use": "No",
"Released": "2026-06-04",
"Good for": "Good for: Many languages; creators may monetize content."
},
{
"Model": "[Chatterbox (Multilingual V3)](https://huggingface.co/ResembleAI/chatterbox)",
"Maker": "Resemble AI",
"Needs": "3.5 GB",
"Quality": "#10 of 12 (Artificial Analysis Speech Arena)",
"Business use": "OK",
"Released": "2025-04-24",
"Good for": "Good for: Free voice cloning with commercial rights."
},
{
"Model": "[VibeVoice 1.5B](https://huggingface.co/microsoft/VibeVoice-1.5B)",
"Maker": "Microsoft",
"Needs": "6.5 GB",
"Quality": "#11 of 12 (Artificial Analysis Speech Arena)",
"Business use": "check the licence",
"Released": "2025-08-25",
"Good for": "Good for: Long podcasts with up to four speakers."
},
{
"Model": "[XTTS v2](https://huggingface.co/coqui/XTTS-v2)",
"Maker": "Coqui",
"Needs": "2.5 GB",
"Quality": "#12 of 12 (Artificial Analysis Speech Arena)",
"Business use": "No",
"Released": "2023-10-31",
"Good for": "Good for: Classic cloning from a 6-second clip."
},
{
"Model": "[Phonon (Gradium)](https://huggingface.co/gradium/phonon-31m-en)",
"Maker": "Gradium",
"Needs": "0.121 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-10-21",
"Good for": "Good for: A tiny voice for a local assistant on a Mac, one Ollama command."
},
{
"Model": "[IndexTTS-2.5](https://huggingface.co/IndexTeam/IndexTTS-2.5) (not on Mac)",
"Maker": "bilibili Index Team",
"Needs": "6 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-08-10",
"Good for": "Good for: Cloning with emotion control."
},
{
"Model": "[VoxCPM2](https://huggingface.co/openbmb/VoxCPM2)",
"Maker": "OpenBMB",
"Needs": "4 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-04-03",
"Good for": "Good for: 48 kHz cloning, free for commercial use."
},
{
"Model": "[OmniVoice](https://huggingface.co/k2-fsa/OmniVoice)",
"Maker": "k2-fsa",
"Needs": "2 GB",
"Quality": "unknown",
"Business use": "No",
"Released": "2026-03-30",
"Good for": "Good for: Rare languages and fast cloning."
},
{
"Model": "[Kitten TTS 0.8 nano](https://huggingface.co/KittenML/kitten-tts-nano-0.8-fp32)",
"Maker": "KittenML",
"Needs": "0.1 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-02-19",
"Good for": "Good for: Tiny CPU-only voice for any machine."
},
{
"Model": "[Qwen3-TTS 1.7B](https://huggingface.co/Qwen/Qwen3-TTS-12Hz-1.7B-Base)",
"Maker": "Alibaba Qwen",
"Needs": "4 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-01-21",
"Good for": "Good for: Cloning, preset voices and voice design."
},
{
"Model": "[Fun-CosyVoice 3 0.5B](https://huggingface.co/FunAudioLLM/Fun-CosyVoice3-0.5B-2512)",
"Maker": "Alibaba FunAudioLLM",
"Needs": "2 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2025-12-11",
"Good for": "Good for: Chinese dialects and low-latency streaming."
},
{
"Model": "[VibeVoice Realtime 0.5B](https://huggingface.co/microsoft/VibeVoice-Realtime-0.5B)",
"Maker": "Microsoft",
"Needs": "1.5 GB",
"Quality": "unknown",
"Business use": "check the licence",
"Released": "2025-12-04",
"Good for": "Good for: Streaming speech as text arrives."
},
{
"Model": "[Dia 1.6B](https://huggingface.co/nari-labs/Dia-1.6B)",
"Maker": "Nari Labs",
"Needs": "10 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2025-04-20",
"Good for": "Good for: Two-speaker dialogue with laughs and sighs."
},
{
"Model": "[F5-TTS v1](https://huggingface.co/SWivid/F5-TTS)",
"Maker": "SWivid",
"Needs": "2 GB",
"Quality": "unknown",
"Business use": "No",
"Released": "2025-03-12",
"Good for": "Good for: Simple zero-shot cloning for hobby projects."
}
]
}
]
},
{
"heading": "Ears (21)",
"level": 3,
"text": [
"Speech in: turns talk into text."
],
"items": [],
"tables": [
{
"columns": [
"Model",
"Maker",
"Needs",
"Quality",
"Business use",
"Released",
"Good for"
],
"rows": [
{
"Model": "[Granite Speech 5.0 2B](https://huggingface.co/ibm-granite/granite-speech-5.0-2b)",
"Maker": "IBM",
"Needs": "6 GB",
"Quality": "#1 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2026-10-20",
"Good for": "Good for: Private transcription and translation on a Mac, business-friendly license."
},
{
"Model": "[Qwen3-ASR 1.7B](https://huggingface.co/Qwen/Qwen3-ASR-1.7B)",
"Maker": "Alibaba Qwen",
"Needs": "3 GB",
"Quality": "#2 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2026-01-28",
"Good for": "Good for: Most accurate open transcription right now."
},
{
"Model": "[Cohere Transcribe (03-2026)](https://huggingface.co/CohereLabs/cohere-transcribe-03-2026)",
"Maker": "Cohere",
"Needs": "5 GB",
"Quality": "#3 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2026-03-26",
"Good for": "Good for: Accurate dictation in 14 major languages."
},
{
"Model": "[Canary-Qwen 2.5B](https://huggingface.co/nvidia/canary-qwen-2.5b)",
"Maker": "NVIDIA",
"Needs": "6 GB",
"Quality": "#4 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2025-07-17",
"Good for": "Good for: High-accuracy English on an NVIDIA GPU."
},
{
"Model": "[Granite Speech 4.1 2B](https://huggingface.co/ibm-granite/granite-speech-4.1-2b)",
"Maker": "IBM",
"Needs": "6 GB",
"Quality": "#5 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2026-04-29",
"Good for": "Good for: Transcribe and translate, business-friendly license."
},
{
"Model": "[Parakeet TDT 0.6B v2](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v2)",
"Maker": "NVIDIA",
"Needs": "3 GB",
"Quality": "#6 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2025-05-01",
"Good for": "Good for: Very fast English with word timestamps."
},
{
"Model": "[MOSS-Transcribe-Diarize 0.9B](https://huggingface.co/OpenMOSS-Team/MOSS-Transcribe-Diarize)",
"Maker": "OpenMOSS",
"Needs": "2.5 GB",
"Quality": "#7 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2026-07-09",
"Good for": "Good for: Meetings: who said what, and when."
},
{
"Model": "[Parakeet TDT 0.6B v3](https://huggingface.co/nvidia/parakeet-tdt-0.6b-v3)",
"Maker": "NVIDIA",
"Needs": "3 GB",
"Quality": "#8 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2025-08-14",
"Good for": "Good for: Fast offline dictation on a Mac."
},
{
"Model": "[Qwen3-ASR 0.6B](https://huggingface.co/Qwen/Qwen3-ASR-0.6B)",
"Maker": "Alibaba Qwen",
"Needs": "1 GB",
"Quality": "#9 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2026-01-28",
"Good for": "Good for: Small multilingual model for lighter machines."
},
{
"Model": "[Granite Speech 5.0 470M TurboCTC](https://huggingface.co/ibm-granite/granite-speech-5.0-470m-turboctc)",
"Maker": "IBM",
"Needs": "1 GB",
"Quality": "#10 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2026-08-25",
"Good for": "Good for: Extremely fast English on laptops and phones."
},
{
"Model": "[Distil-Whisper large-v3.5](https://huggingface.co/distil-whisper/distil-large-v3.5)",
"Maker": "Hugging Face (distil-whisper)",
"Needs": "2 GB",
"Quality": "#11 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2024-12-05",
"Good for": "Good for: Faster drop-in Whisper for English."
},
{
"Model": "[Whisper large-v3](https://huggingface.co/openai/whisper-large-v3)",
"Maker": "OpenAI",
"Needs": "3.9 GB",
"Quality": "#12 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2023-11-07",
"Good for": "Good for: The safe multilingual default."
},
{
"Model": "[Voxtral Mini 3B (2507)](https://huggingface.co/mistralai/Voxtral-Mini-3B-2507)",
"Maker": "Mistral AI",
"Needs": "9.5 GB",
"Quality": "#13 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "unknown",
"Good for": "Good for: Transcribe, then ask questions about audio."
},
{
"Model": "[Canary 1B v2](https://huggingface.co/nvidia/canary-1b-v2)",
"Maker": "NVIDIA",
"Needs": "4.5 GB",
"Quality": "#14 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2025-08-14",
"Good for": "Good for: European languages plus translation to English."
},
{
"Model": "[Whisper large-v3-turbo](https://huggingface.co/openai/whisper-large-v3-turbo)",
"Maker": "OpenAI",
"Needs": "2 GB",
"Quality": "#15 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2024-10-01",
"Good for": "Good for: Fast multilingual dictation, runs everywhere."
},
{
"Model": "[VibeVoice-ASR](https://huggingface.co/microsoft/VibeVoice-ASR)",
"Maker": "Microsoft",
"Needs": "11 GB",
"Quality": "#16 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2026-01-21",
"Good for": "Good for: Hour-long recordings with speakers and timestamps."
},
{
"Model": "[Voxtral Mini 4B Realtime (2602)](https://huggingface.co/mistralai/Voxtral-Mini-4B-Realtime-2602)",
"Maker": "Mistral AI",
"Needs": "4 GB",
"Quality": "#17 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2026-02-04",
"Good for": "Good for: Live captions and voice assistants."
},
{
"Model": "[Nemotron 3.5 ASR Streaming 0.6B](https://huggingface.co/nvidia/nemotron-3.5-asr-streaming-0.6b)",
"Maker": "NVIDIA",
"Needs": "1.5 GB",
"Quality": "#18 of 18 (Open ASR Leaderboard (English))",
"Business use": "OK",
"Released": "2026-06-04",
"Good for": "Good for: Low-latency multilingual streaming for voice agents."
},
{
"Model": "[Moonshine Streaming Medium](https://huggingface.co/moonshine-ai/moonshine-streaming-medium)",
"Maker": "Moonshine AI (Useful Sensors)",
"Needs": "1 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2026-01-06",
"Good for": "Good for: Live English captions on small devices."
},
{
"Model": "[Whisper small](https://huggingface.co/openai/whisper-small)",
"Maker": "OpenAI",
"Needs": "0.9 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2022-09-26",
"Good for": "Good for: Light dictation on older or low-RAM machines."
},
{
"Model": "[Whisper base.en](https://huggingface.co/openai/whisper-base.en)",
"Maker": "OpenAI",
"Needs": "0.4 GB",
"Quality": "unknown",
"Business use": "OK",
"Released": "2022-09-26",
"Good for": "Good for: Small, quick English dictation on low-memory machines."
}
]
}
]
},
{
"heading": "Open-source tools",
"level": 3,
"text": [
"237 free, open-source tools in 19 aisles (catalog.json; 4 cloud-only or closed tools are left out of the shop). Add one to My setup and checkout puts it in MY-CART.md with its install lines from its README, which we have not tested. Memory: Hermes Agent remembers you already, so those tools are not sold as add-ons. Coding agents, General & personal agents and Chat UIs are other apps like Hermes Agent: shown to compare, not added."
],
"items": [
"**Runtimes & inference** (19): Runs open models on your own hardware. Most stars: Ollama (in every setup already) (184,859 stars), llama.cpp (132,426 stars), PyTorch (104,960 stars), vLLM (93,863 stars), LocalAI (49,849 stars). All: https://insidethemachine.ai/setups/c/runtimes",
"**Model hubs** (9): Where you find and download weights. Most stars: Transformers (167,719 stars), Unsloth notebooks (5,738 stars), Hugging Face Hub, Ollama Library, LM Studio model catalog. All: https://insidethemachine.ai/setups/c/hubs",
"**Models from the channel** (6): Featured on Inside the Machine; check the fit before you download. Most stars: Qwen-Image-2.1-Turbo, Mellum2.1 12B-A2.5B, EmbeddingGemma 2, Kolibri-1, AstaBrief 8B. All: https://insidethemachine.ai/setups/c/models",
"**Coding agents** (13, other apps like Hermes Agent): Turns a model into something that edits your code. Most stars: OpenCode (213,869 stars), Codex CLI (129,383 stars), Pi (pi-mono) (115,690 stars), OpenHands (90,983 stars), Cline (70,365 stars). All: https://insidethemachine.ai/setups/c/coding",
"**General & personal agents** (14, other apps like Hermes Agent): Lets a model take actions for you, not just chat. Most stars: OpenClaw (392,108 stars), Hermes Agent (in every setup already) (254,350 stars), AutoGPT (187,599 stars), Browser Use (118,088 stars), DeerFlow (83,911 stars). All: https://insidethemachine.ai/setups/c/general",
"**Agent frameworks** (14): Building blocks for your own agents. Most stars: AutoGen (61,426 stars), LiteLLM (61,329 stars), CrewAI (59,740 stars), LangGraph (43,319 stars), Agno (42,754 stars). All: https://insidethemachine.ai/setups/c/frameworks",
"**Memory** (11, already in Hermes Agent): Lets your agent remember across sessions. Most stars: claude-mem (100,993 stars), Mem0 (67,235 stars), Cognee (32,356 stars), Graphiti (31,795 stars), Supermemory (31,236 stars). All: https://insidethemachine.ai/setups/c/memory",
"**RAG & document parsing** (14): Feeds your own documents to the model. Most stars: MarkItDown (190,405 stars), RAGFlow (92,283 stars), PaddleOCR (91,178 stars), MinerU (81,815 stars), Docling (68,936 stars). All: https://insidethemachine.ai/setups/c/rag",
"**Routing & gateways** (10): Sends each request to the right model. Most stars: New API (50,036 stars), One API (37,168 stars), Portkey Gateway (13,194 stars), Bifrost (8,833 stars), Plano (7,090 stars). All: https://insidethemachine.ai/setups/c/routing",
"**MCP servers & tools** (14): Gives the agent hands: web, files, apps. Most stars: n8n (207,507 stars), Firecrawl (191,861 stars), Awesome MCP Servers (96,394 stars), MCP Servers (official) (91,287 stars), Crawl4AI (85,617 stars). All: https://insidethemachine.ai/setups/c/mcp",
"**Chat UIs & front-ends** (14, other apps like Hermes Agent): A friendly window onto your local models. Most stars: Open WebUI (155,901 stars), NextChat (88,853 stars), LobeChat (83,250 stars), Cherry Studio (52,668 stars), text-generation-webui (47,730 stars). All: https://insidethemachine.ai/setups/c/ui",
"**Voice: speech-to-text & TTS** (14): Talk to it and hear it answer, offline. Most stars: Whisper (110,561 stars), whisper.cpp (54,418 stars), ChatTTS (39,909 stars), Handy (33,870 stars), Fish Speech (33,167 stars). All: https://insidethemachine.ai/setups/c/voice",
"**Image & video generation** (14): Make images and video on your own machine. Most stars: Stable Diffusion WebUI (165,059 stars), ComfyUI (137,692 stars), Fooocus (53,435 stars), Diffusers (34,726 stars), Open-Sora (29,873 stars). All: https://insidethemachine.ai/setups/c/image",
"**Fine-tuning** (14): Teach a model your data and style. Most stars: Unsloth (78,064 stars), LLaMA-Factory (75,495 stars), nanoGPT (63,835 stars), nanochat (58,660 stars), Open R1 (26,481 stars). All: https://insidethemachine.ai/setups/c/finetune",
"**Quantization** (11): Shrinks models so bigger ones fit in memory. Most stars: bitsandbytes (8,527 stars), Core ML Tools (5,435 stars), TensorRT Model Optimizer (5,326 stars), LLM Compressor (3,878 stars), Optimum (3,503 stars). All: https://insidethemachine.ai/setups/c/quant",
"**Evals & benchmarks** (14): Measures how close you are to the frontier. Most stars: promptfoo (26,019 stars), OpenAI Evals (19,607 stars), DeepEval (18,890 stars), Ragas (16,034 stars), LM Evaluation Harness (14,234 stars). All: https://insidethemachine.ai/setups/c/evals",
"**Scaffolds & templates** (11): Reusable instructions that make agents behave. Most stars: Spec Kit (141,412 stars), OpenSpec (72,037 stars), Awesome Claude Code (55,762 stars), BMAD Method (54,187 stars), Awesome Cursor Rules (40,912 stars). All: https://insidethemachine.ai/setups/c/scaffolds",
"**Observability** (12): See what your agent did and what it cost. Most stars: Langfuse (35,875 stars), MLflow (28,399 stars), Opik (22,627 stars), Phoenix (11,858 stars), Evidently (7,991 stars). All: https://insidethemachine.ai/setups/c/obs",
"**Hardware & system utilities** (9): Know what your machine can run, and watch it work. Most stars: Apple container (50,702 stars), Stats (42,577 stars), llmfit (37,950 stars), nvtop (11,089 stars), nvitop (7,210 stars). All: https://insidethemachine.ai/setups/c/hw"
],
"tables": []
},
{
"heading": "Skills (27)",
"level": 3,
"text": [
"A skill is a folder with a SKILL.md file: instructions that teach an agent one job. Hermes Agent, our app, loads skills itself (hermes skills). Each pack below is a GitHub repo with many skills: add it to My setup and MY-CART.md tells the agent to install only the skills the person needs. Install lines come from the Hermes Agent skills docs and the npx skills docs (vercel-labs/skills), read 2026-10-21; we have not tested them. Skill counts: SKILL.md files in the repo on GitHub. All: https://insidethemachine.ai/setups/c/skills"
],
"items": [],
"tables": []
},
{
"heading": "For everyone (5)",
"level": 4,
"text": [
"Office files, Hugging Face and science. Good first skills."
],
"items": [],
"tables": [
{
"columns": [
"Pack",
"What",
"Skills",
"Stars",
"Licence",
"Needs"
],
"rows": [
{
"Pack": "[Anthropic Skills](https://insidethemachine.ai/setups/anthropic-skills)",
"What": "Make PDFs, Word, Excel and slide files; write new skills.",
"Skills": "20",
"Stars": "182,747",
"Licence": "no licence file on GitHub: check each skill",
"Needs": "no paid service"
},
{
"Pack": "[Scientific Agent Skills](https://insidethemachine.ai/setups/scientific-agent-skills)",
"What": "177 science skills: biology, chemistry, medicine.",
"Skills": "177",
"Stars": "53,569",
"Licence": "MIT",
"Needs": "no paid service"
},
{
"Pack": "[OpenAI Skills](https://insidethemachine.ai/setups/openai-skills)",
"What": "OpenAI's skill catalog, made for Codex.",
"Skills": "44",
"Stars": "28,226",
"Licence": "no licence file on GitHub: check each skill",
"Needs": "no paid service"
},
{
"Pack": "[Hugging Face Skills](https://insidethemachine.ai/setups/huggingface-skills)",
"What": "Find, run, train and share models on Hugging Face.",
"Skills": "26",
"Stars": "11,194",
"Licence": "Apache-2.0",
"Needs": "no paid service"
},
{
"Pack": "[OpenScience](https://insidethemachine.ai/setups/openscience)",
"What": "A science workbench with hundreds of research skills.",
"Skills": "373",
"Stars": "4,189",
"Licence": "Apache-2.0",
"Needs": "no paid service"
}
]
}
]
},
{
"heading": "Content creation (8)",
"level": 4,
"text": [
"Make images, video and voice, and write better prompts for them. All but HyperFrames call a paid cloud service: you need an account there, and each image or video costs money."
],
"items": [],
"tables": [
{
"columns": [
"Pack",
"What",
"Skills",
"Stars",
"Licence",
"Needs"
],
"rows": [
{
"Pack": "[HyperFrames](https://insidethemachine.ai/setups/hyperframes)",
"What": "Write HTML animation and render it to an MP4 video.",
"Skills": "7",
"Stars": "69,168",
"Licence": "Apache-2.0",
"Needs": "no paid service"
},
{
"Pack": "[Seedance 2.0 Skills](https://insidethemachine.ai/setups/seedance-skills)",
"What": "Seedance 2.0 video prompts: camera, light, sound, fixes.",
"Skills": "29",
"Stars": "7,882",
"Licence": "MIT",
"Needs": "Writes prompts only. Making the video needs a paid Seedance service."
},
{
"Pack": "[Higgsfield Skills](https://insidethemachine.ai/setups/higgsfield-skills)",
"What": "Images and video with GPT Image, Seedance, Kling, Veo.",
"Skills": "8",
"Stars": "1,354",
"Licence": "MIT",
"Needs": "Needs a Higgsfield account with credits. Each image or video costs money."
},
{
"Pack": "[HeyGen Skills](https://insidethemachine.ai/setups/heygen-skills)",
"What": "Talking avatar videos and dubbing with HeyGen.",
"Skills": "3",
"Stars": "509",
"Licence": "MIT",
"Needs": "Needs a HeyGen account and an API key (or its sign-in). It costs money."
},
{
"Pack": "[ElevenLabs Skills](https://insidethemachine.ai/setups/elevenlabs-skills)",
"What": "Voice, music and sound effects with ElevenLabs.",
"Skills": "12",
"Stars": "500",
"Licence": "MIT",
"Needs": "Needs an ElevenLabs account and an API key. It costs money."
},
{
"Pack": "[fal Skills](https://insidethemachine.ai/setups/fal-skills)",
"What": "Pick, prompt and chain fal.ai image and video models.",
"Skills": "17",
"Stars": "254",
"Licence": "no licence file on GitHub (the README says MIT): check each skill",
"Needs": "Needs a fal.ai account and an API key. Each run costs money."
},
{
"Pack": "[Replicate Skills](https://insidethemachine.ai/setups/replicate-skills)",
"What": "Find, compare and prompt image and video models.",
"Skills": "7",
"Stars": "136",
"Licence": "Apache-2.0",
"Needs": "Needs a Replicate account and an API token. Each run costs money."
},
{
"Pack": "[Black Forest Labs Skills](https://insidethemachine.ai/setups/bfl-skills)",
"What": "FLUX image and video prompts from the FLUX makers.",
"Skills": "10",
"Stars": "134",
"Licence": "MIT",
"Needs": "Needs a Black Forest Labs account and an API key. Each run costs money."
}
]
}
]
},
{
"heading": "For coding (5)",
"level": 4,
"text": [
"Make a coding agent plan, test, review and find bugs."
],
"items": [],
"tables": [
{
"columns": [
"Pack",
"What",
"Skills",
"Stars",
"Licence",
"Needs"
],
"rows": [
{
"Pack": "[Superpowers](https://insidethemachine.ai/setups/superpowers)",
"What": "Makes coding agents plan, test and review their work.",
"Skills": "15",
"Stars": "303,133",
"Licence": "MIT",
"Needs": "no paid service"
},
{
"Pack": "[Matt Pocock's Skills](https://insidethemachine.ai/setups/mattpocock-skills)",
"What": "Skills for real engineering: specs, TDD, code review.",
"Skills": "38",
"Stars": "299,798",
"Licence": "MIT",
"Needs": "no paid service"
},
{
"Pack": "[gstack](https://insidethemachine.ai/setups/gstack)",
"What": "Garry Tan's Claude Code setup: CEO, designer, QA roles.",
"Skills": "57",
"Stars": "137,200",
"Licence": "MIT",
"Needs": "no paid service"
},
{
"Pack": "[Vercel Agent Skills](https://insidethemachine.ai/setups/vercel-agent-skills)",
"What": "React, Next.js and web design rules from Vercel.",
"Skills": "9",
"Stars": "32,771",
"Licence": "no licence file on GitHub: check each skill",
"Needs": "no paid service"
},
{
"Pack": "[Trail of Bits Skills](https://insidethemachine.ai/setups/trailofbits-skills)",
"What": "Security skills: find bugs and audit code.",
"Skills": "97",
"Stars": "7,687",
"Licence": "CC-BY-SA-4.0",
"Needs": "no paid service"
}
]
}
]
},
{
"heading": "For building on one platform (9)",
"level": 4,
"text": [
"Each pack teaches your agent one company's product. Most of those products are cloud services."
],
"items": [],
"tables": [
{
"columns": [
"Pack",
"What",
"Skills",
"Stars",
"Licence",
"Needs"
],
"rows": [
{
"Pack": "[Remotion Skills](https://insidethemachine.ai/setups/remotion-skills)",
"What": "Skills for making videos in code with Remotion.",
"Skills": "12",
"Stars": "5,484",
"Licence": "no licence file on GitHub: check each skill",
"Needs": "an account with that company's product (many are paid cloud services)"
},
{
"Pack": "[Gemini Skills](https://insidethemachine.ai/setups/gemini-skills)",
"What": "Skills for building with Google's Gemini API.",
"Skills": "3",
"Stars": "4,317",
"Licence": "Apache-2.0",
"Needs": "an account with that company's product (many are paid cloud services)"
},
{
"Pack": "[NVIDIA Skills](https://insidethemachine.ai/setups/nvidia-skills)",
"What": "NVIDIA skills: CUDA, robotics, simulation, RAG.",
"Skills": "424",
"Stars": "3,683",
"Licence": "Apache-2.0",
"Needs": "an account with that company's product (many are paid cloud services)"
},
{
"Pack": "[Microsoft Skills](https://insidethemachine.ai/setups/microsoft-skills)",
"What": "Skills for Azure and Microsoft SDKs.",
"Skills": "212",
"Stars": "3,137",
"Licence": "MIT",
"Needs": "an account with that company's product (many are paid cloud services)"
},
{
"Pack": "[Cloudflare Skills](https://insidethemachine.ai/setups/cloudflare-skills)",
"What": "Skills for building on Cloudflare.",
"Skills": "19",
"Stars": "3,133",
"Licence": "Apache-2.0",
"Needs": "an account with that company's product (many are paid cloud services)"
},
{
"Pack": "[Expo Skills](https://insidethemachine.ai/setups/expo-skills)",
"What": "Skills for Expo and React Native apps.",
"Skills": "23",
"Stars": "2,973",
"Licence": "MIT",
"Needs": "an account with that company's product (many are paid cloud services)"
},
{
"Pack": "[Supabase Skills](https://insidethemachine.ai/setups/supabase-skills)",
"What": "Skills for Supabase and Postgres.",
"Skills": "2",
"Stars": "2,902",
"Licence": "MIT",
"Needs": "an account with that company's product (many are paid cloud services)"
},
{
"Pack": "[Callstack Skills](https://insidethemachine.ai/setups/callstack-skills)",
"What": "React Native skills from Callstack.",
"Skills": "9",
"Stars": "1,679",
"Licence": "MIT",
"Needs": "an account with that company's product (many are paid cloud services)"
},
{
"Pack": "[Sentry Skills](https://insidethemachine.ai/setups/sentry-skills)",
"What": "The Sentry team's skills: reviews, commits, bug hunts.",
"Skills": "29",
"Stars": "1,080",
"Licence": "Apache-2.0",
"Needs": "an account with that company's product (many are paid cloud services)"
}
]
}
]
},
{
"heading": "The weekly email: The Local Drop",
"level": 2,
"text": [
"The Local Drop · free, weekly. Every new AI you can run at home. One email a week. New models, tools and hardware, and what each one needs to run. Sign up on this page: https://insidethemachine.ai/#report. Gave your email at checkout? You are already on the list."
],
"items": [],
"tables": []
},
{
"heading": "Where next",
"level": 2,
"text": [],
"items": [
"Learn, with How good is local AI? (the smartest model for each memory size, as a share of the best paid AI, Claude Opus 5.5): https://insidethemachine.ai/learn.md, https://insidethemachine.ai/best.md. The plan for our own Gap Score: https://insidethemachine.ai/gap-score.md",
"The directory: The shop https://insidethemachine.ai/shop · Skills https://insidethemachine.ai/skills · Models (130 open text models) https://insidethemachine.ai/models · Agent apps (36 apps) https://insidethemachine.ai/harnesses · Machines (56 machines) https://insidethemachine.ai/machines · Tools (268 tools) https://insidethemachine.ai/setups · Recipes (6 recipes) https://insidethemachine.ai/do · Scan my machine https://insidethemachine.ai/my · Best model for your RAM https://insidethemachine.ai/best · Compare apps https://insidethemachine.ai/compare · New this week https://insidethemachine.ai/changes"
],
"tables": []
}
]
}