# Best local LLM for 48GB RAM or VRAM

URL: https://insidethemachine.ai/best/local-llm-for-48gb-ram
Data: models 2026-10-15, machines 2026-10-15. Source of scores: Artificial Analysis Intelligence Index (AA), LMArena. Not our own tests yet.
Usable memory assumed: 36 GB (75% of 48). Rule: min_memory_gb = weights_gb_4bit × 1.15 + 2. weights_gb_4bit = Q4_K_M file (Ollama, else HF GGUF) → MLX 4-bit (HF) → Ollama MLX/NVFP4 tag → Ollama default tag → params_total_b × 0.6 (flagged weights_estimated). +15% ≈ KV cache at ~8–16K context; +2GB runtime/process overhead. Long contexts (128K+) need much more.
Models that fit: 76 of 124.

## Top picks

| model | maker | AA index | needs (4-bit) | needs (8-bit) | run |
|---|---|---|---|---|---|
| Qwen3.8 27B | Alibaba Qwen | 33.7 | 22.7 GB | 36.5 GB | `ollama run qwen3.8:27b` |
| K2 Horizon MoVA 36B-A4B | IFM (MBZUAI) | 25.3 | 28.1 GB | 47.8 GB | MLX only |
| Qwen3.5 27B | Alibaba Qwen | 22.9 | 21.5 GB | 36.5 GB | `ollama run qwen3.5:27b` |
| Qwen3.6 27B | Alibaba Qwen | 21.4 | 21.5 GB | 36.5 GB | `ollama run qwen3.6:27b` |
| Nemotron 3.5 Bolt 74B-A9B | NVIDIA | 21.4 | 30.7 GB | - | `ollama run nemotron-3.5-bolt:74b-a9b` |
| K2 Horizon 7B | IFM (MBZUAI) | 20.6 | 8.6 GB | 13 GB | MLX only |

## By job

- Coding: Qwen3.8 27B (22.7 GB, Terminal-Bench 2.1 79.8%) `ollama run qwen3.8:27b`
- Agents (tool use): Qwen3.8 27B (22.7 GB) `ollama run qwen3.8:27b`
- Vision (reads images): Qwen3.8 27B (22.7 GB)

## New at 48 GB (did not fit in 32 GB)

- K2 Horizon MoVA 36B-A4B (28.1 GB, AA 25.3)
- Nemotron 3.5 Bolt 74B-A9B (30.7 GB, AA 21.4)
- Qwen3.5 35B-A3B (27.3 GB, AA 19.3)
- Qwen3.6 35B-A3B (29.6 GB, AA 18.2)
- Gemma 4 31B (25 GB, AA 14.7)

## Machines with 48 GB

| machine | type | usable GB | models that fit | best fit | from (approx) |
|---|---|---|---|---|---|
| MacBook Pro 14"/16" (M5 Pro) | Mac laptop | 36 | 76 | Qwen3.8 27B | $2,499 (2026-10-15) |
| MacBook Pro 14"/16" (M5 Max) | Mac laptop | 36 | 76 | Qwen3.8 27B | $4,099 (2026-10-15) |
| Mac mini (M5 Pro) | Mac desktop | 36 | 76 | Qwen3.8 27B | $1,699 (2026-10-15) |
| Mac Studio (M5 Max) | Mac desktop | 36 | 76 | Qwen3.8 27B | $2,499 (2026-10-15) |

## Questions

**What is the best local LLM for 48 GB of RAM?**
Qwen3.8 27B by Alibaba Qwen. It scores 33.7 on the Artificial Analysis Intelligence Index and needs about 22.7 GB at 4-bit. That is about 59% of Claude Opus 5.5 (57.6), the top paid model. Data as of 2026-10-15.

**How many local models fit in 48 GB?**
76 of the 124 text models in our data fit. We assume about 36 GB is usable for the model (75% of 48 GB), and a model needs its 4-bit file size × 1.15 + 2 GB.

**What is the biggest model that runs in 48 GB?**
Nemotron 3.5 Bolt 74B-A9B, 74B (9B active) parameters, needs about 30.7 GB at 4-bit. Bigger is not always smarter: Qwen3.8 27B scores 33.7; Nemotron 3.5 Bolt 74B-A9B scores 21.4.

**What is the best coding model for 48 GB?**
Qwen3.8 27B (needs 22.7 GB, Terminal-Bench 2.1 79.8%). Run it in a coding app like OpenCode or Cline. See https://insidethemachine.ai/best/local-coding-model.

**Does 48 GB of VRAM work the same as 48 GB of RAM?**
Close. On a Mac the memory is shared, and about 75% goes to the model. On a PC the graphics card's VRAM is what counts; we estimate VRAM minus 1 GB is usable, so a 48 GB card fits a little more than 48 GB of Mac memory. Models that spill out of VRAM into normal RAM run much slower.

More: https://insidethemachine.ai/models?tier=48 · https://insidethemachine.ai/machines?tier=48 · https://insidethemachine.ai/best