Best local LLMs for 64 GB unified memory
Every catalogued model that fits entirely in 64 GB, with no layers moved to system RAM. Speeds are estimated on the Apple M1 Max.
- Largest popular model that fitsQwen3-Coder-Next80B params · Q4_K_M · needs 50.0 GB~76 tok/s Faster than you readOpen in the calculator →
- Best fast pickQwen3-Next-80B-A3B-Instruct81B params · Q4_K_M · needs 50.0 GB~76 tok/s Faster than you readOpen in the calculator →
- Most downloaded that fitsQwen3.8-27B28B params · Q4_K_M · needs 18.6 GB~22 tok/s About reading paceOpen in the calculator →
“Best” is not a quality ranking. The picks are the largest widely downloaded model that fits, the largest that still answers at 30 tokens a second or more, and the most downloaded; the list puts current, widely downloaded releases first. Speeds are estimates from memory bandwidth, not benchmarks.
Is 64 GB of unified memory enough for a local LLM?
For models under 36B parameters, yes: all 211 in the catalogue fit entirely at Q4_K_M, the format most people download. In the 36B and larger band, 19 of 51 fit. 183 of the 233 that fit answer at 30 tokens a second or more on the Apple M1 Max, which is faster than most people read. A longer conversation needs more memory than the 8,192 tokens counted here, and a smaller format than Q4_K_M needs less; the calculator sizes either.
| Model size | Fit in 64 GB | For example |
|---|---|---|
| Under 4B parameters | 66 of 66 | Qwen3.5-2B · needs 2.32 GB |
| 4B to 9B parameters | 58 of 58 | Qwen3.5-4B · needs 3.98 GB |
| 9B to 16B parameters | 21 of 21 | Qwen3.5-9B · needs 7.05 GB |
| 16B to 36B parameters | 66 of 66 | Qwen3.8-27B · needs 18.6 GB |
| 36B and larger parameters | 19 of 51 | Qwen3-Coder-Next · needs 50.0 GB |
Models that fit in 64 GB
| Model | Parameters | Needs | Spare | Decode | Answer |
|---|---|---|---|---|---|
| Qwen3.8-27B Qwen · Q4_K_M | 28B | 18.6 GB | 45.4 GB | ~22 tok/s | Details |
| gemma-4-26B-A4B-it Google · Q4_K_M | 26B | 16.8 GB | 47.2 GB | ~78 tok/s | Details |
| gemma-4-31B-it Google · Q4_K_M | 31B | 22.2 GB | 41.8 GB | ~16 tok/s | Details |
| Qwen3.5-9B Qwen · Q4_K_M | 9.7B | 7.05 GB | 57.0 GB | ~54 tok/s | Details |
| Qwen3.5-4B Qwen · Q4_K_M | 4.7B | 3.98 GB | 60.0 GB | ~78 tok/s | Details |
| Qwen3.6-35B-A3B Qwen · Q4_K_M | 36B | 23.1 GB | 40.9 GB | ~94 tok/s | Details |
| gemma-4-12B-it Google · Q4_K_M | 12B | 9.54 GB | 54.5 GB | ~36 tok/s | Details |
| Qwen3.6-27B Qwen · Q4_K_M | 28B | 18.6 GB | 45.4 GB | ~22 tok/s | Details |
| Qwen3.5-2B Qwen · Q4_K_M | 2.3B | 2.32 GB | 61.7 GB | ~111 tok/s | Details |
| gemma-4-E4B-it Google · Q4_K_M | 8.0B | 5.86 GB | 58.1 GB | ~73 tok/s | Details |
| NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16 NVIDIA · Q4_K_M | 32B | 20.3 GB | 43.7 GB | ~17 tok/s | Details |
| NVIDIA-Nemotron-3-Nano-4B-BF16 NVIDIA · Q4_K_M | 4.0B | 3.46 GB | 60.5 GB | ~78 tok/s | Details |
| gemma-4-E2B-it Google · Q4_K_M | 5.1B | 4.02 GB | 60.0 GB | ~114 tok/s | Details |
| Qwen3.5-0.8B Qwen · Q4_K_M | 873M | 1.46 GB | 62.5 GB | ~139 tok/s | Details |
| Qwen3-VL-8B-Instruct Qwen · Q4_K_M | 8.8B | 7.04 GB | 57.0 GB | ~47 tok/s | Details |
| Qwen3.5-27B Qwen · Q4_K_M | 28B | 18.6 GB | 45.4 GB | ~22 tok/s | Details |
| North-Micro-Vision-Instruct CohereLabs · Q4_K_M | 2.5B | 2.99 GB | 61.0 GB | ~94 tok/s | Details |
| Qwen3.5-35B-A3B Qwen · Q4_K_M | 36B | 23.1 GB | 40.9 GB | ~94 tok/s | Details |
| granite-4.2-8b IBM · Q4_K_M | 8.8B | 7.49 GB | 56.5 GB | ~44 tok/s | Details |
| GLM-4.7-Flash zai-org · Q4_K_M | 31B | 20.4 GB | 43.6 GB | ~17 tok/s | Details |
| granite-4.1-3b IBM · Q4_K_M | 3.4B | 3.72 GB | 60.3 GB | ~76 tok/s | Details |
| LFM2.5-2.6B LiquidAI · Q4_K_M | 2.7B | 2.61 GB | 61.4 GB | ~97 tok/s | Details |
| Qwen3-VL-4B-Instruct Qwen · Q4_K_M | 4.4B | 4.72 GB | 59.3 GB | ~64 tok/s | Details |
| granite-4.1-30b IBM · Q4_K_M | 29B | 20.4 GB | 43.6 GB | ~17 tok/s | Details |
| Qwen3-VL-2B-Instruct Qwen · Q4_K_M | 2.1B | 3.05 GB | 61.0 GB | ~90 tok/s | Details |
| granite-4.2-3b IBM · Q4_K_M | 3.7B | 3.72 GB | 60.3 GB | ~76 tok/s | Details |
| LFM2.5-230M LiquidAI · Q4_K_M | 230M | 1.05 GB | 62.9 GB | ~158 tok/s | Details |
| Qwen3-0.6B Qwen · Q4_K_M | 752M | 2.22 GB | 61.8 GB | ~108 tok/s | Details |
| Qwen3-Coder-Next Qwen · Q4_K_M | 80B | 50.0 GB | 14.0 GB | ~76 tok/s | Details |
| gpt-oss-20b OpenAI · Q4_K_M | 21B | 13.8 GB | 50.2 GB | ~81 tok/s | Details |
| granite-4.2-30b IBM · Q4_K_M | 29B | 20.7 GB | 43.3 GB | ~17 tok/s | Details |
| granite-4.1-8b IBM · Q4_K_M | 8.8B | 7.49 GB | 56.5 GB | ~44 tok/s | Details |
| NVIDIA-Nemotron-3-Nano-30B-A3B-BF16 NVIDIA · Q4_K_M | 32B | 20.3 GB | 43.7 GB | ~17 tok/s | Details |
| LFM2.5-VL-3B LiquidAI · Q4_K_M | 3.1B | 2.85 GB | 61.1 GB | ~97 tok/s | Details |
| MiMo-V2.6-Distill-Qwen-9B XiaomiMiMo · Q4_K_M | 9.4B | 6.90 GB | 57.1 GB | ~54 tok/s | Details |
| Qwen3-4B-Instruct-2507 Qwen · Q4_K_M | 4.0B | 4.72 GB | 59.3 GB | ~64 tok/s | Details |
| Ling-3.0-tiny inclusionAI · Q4_K_M | 7.9B | 5.85 GB | 58.1 GB | ~125 tok/s | Details |
| Qwen3-8B Qwen · Q4_K_M | 8.2B | 7.04 GB | 57.0 GB | ~47 tok/s | Details |
| Olmo-3-7B-Instruct allenai · Q4_K_M | 7.3B | 7.96 GB | 56.0 GB | ~41 tok/s | Details |
| Qwen3-4B Qwen · Q4_K_M | 4.0B | 4.51 GB | 59.5 GB | ~64 tok/s | Details |
3 devices with 64 GB
The same models fit on every one of them. What changes is speed, which follows memory bandwidth: a card with twice the bandwidth decodes roughly twice as fast.
| Device | Bandwidth | Memory | Where to find one |
|---|---|---|---|
| Apple M1 Max speeds on this page | 400 GB/s | 64 GB, unified | Amazon ↗ · eBay (new and used) ↗ |
| Apple M4 Pro | 273 GB/s | 64 GB, unified | Amazon ↗ · eBay (new and used) ↗ |
| Apple M5 Pro | 307 GB/s | 64 GB, unified | Amazon ↗ · eBay (new and used) ↗ |
Which card to buy, at every memory size →
Store links may pay us a commission. They never decide which models are listed — the memory arithmetic does.
What does not fit in 64 GB
Widely downloaded models that need more than 64 GB at Q4_K_M, and the smallest memory size that holds each one entirely. Each link shows what 64 GB can still do with it: a smaller format, or part of the model in system RAM at a lower speed.
| Model | Needs | Fits from | On 64 GB |
|---|---|---|---|
| NVIDIA-Nemotron-3-Super-120B-A12B-BF16 | 76.9 GB | 96 GB | On 64 GB |
| Qwen3.5-122B-A10B | 77.9 GB | 96 GB | On 64 GB |
| gpt-oss-120b | 71.9 GB | 96 GB | On 64 GB |
| Ling-3.0-flash | 78.2 GB | 96 GB | On 64 GB |
| Qwen3.8-Flash-Next | 111.6 GB | 128 GB | On 64 GB |
| DeepSeek-V4-Flash-0731 | 187.5 GB | 192 GB | On 64 GB |
12 more models fit in 96 GB. Best local LLMs for 96 GB →