Arc Pro B70 32G — what LLMs can it run?
72 of 90 indexed models fit comfortably in 32GB at 4K context, each at the highest-quality quant that still leaves headroom.
The short answer for Arc Pro B70 32G
32 GB, 608 GB/s GDDR6. 72 of 90 models in this index fit comfortably at 4K context; 75 load at all.
- Biggest that fits
- Seed-OSS 36B Instruct — 27.8 GB at Q5_K_M — 4.2 GB spare at 4K, so longer context comes out of a thin margin.
- Room to grow
- GLM-4.7-Flash — 19.2 GB at Q4_K_M — under 60% of the card, which leaves 12.8 GB for a long context window or a second process.
- Speed ceiling
- Llama 3.1 8B Instruct at Q4_K_M reads 4.6 GB of weights per token, so 608 GB/s puts a hard ceiling near 132 tok/s. That is arithmetic on two published numbers, not a benchmark — real throughput lands below it. No run on this card has been measured here, so there is nothing to compare the ceiling against.
- The wall you will hit
- Backend coverage, not capacity. At 608 GB/s the card has the memory system for its size class; what limits it is which runtimes reach it. llama.cpp's SYCL backend lists Arc A-series and B580 cards as verified, and Ollama runs them through Vulkan — so GGUF is the format this site recommends here. AWQ, GPTQ and EXL2 are not.
32B · 22
| Model | Quant | Est. VRAM | Headroom | tok/s on RTX 4090 |
|---|---|---|---|---|
| Seed-OSS 36B Instruct36B | Q5_K_M | 27.81 GB | +4.2 GB | — |
| Command R 35B35B | Q4_K_M | 22.86 GB | +9.1 GB | 42Estimated |
| Yi 1.5 34B Chat34B | Q4_K_M | 22.82 GB | +9.2 GB | 40Estimated |
| Qwen3 32B Instruct32B | Q4_K_M | 21.85 GB | +10.1 GB | 42 |
| Qwen2.5 32B Instruct32B | Q4_K_M | 21.69 GB | +10.3 GB | 44 |
| Qwen2.5-Coder 32B Instruct32B | Q4_K_M | 21.69 GB | +10.3 GB | 44Estimated |
| DeepSeek-R1-Distill-Qwen-32B32B | Q4_K_M | 21.69 GB | +10.3 GB | 42Estimated |
| Gemma 4 31B IT31B | Q4_K_M | 21.08 GB | +10.9 GB | — |
| Qwen3 30B-A3B Instruct30B-A3B | Q5_K_M | 23.04 GB | +9 GB | 82 |
| Qwen3-Coder 30B-A3B Instruct30B-A3B | Q5_K_M | 23.04 GB | +9 GB | 80 |
| Qwen3-VL 30B-A3B Instruct30B-A3B | Q4_K_M | 19.73 GB | +12.3 GB | 95Community |
| GLM-4.7-Flash30B-A3B | Q4_K_M | 19.19 GB | +12.8 GB | — |
| Qwen3.8 27B27B | Q4_K_M | 17.47 GB | +14.5 GB | — |
| Gemma 3 27B IT27B | Q4_K_M | 18.37 GB | +13.6 GB | 48Community |
| Gemma 2 27B Instruct27B | Q5_K_M | 21.76 GB | +10.2 GB | 42Estimated |
| Gemma 4 26B-A4B IT26B-A4B | Q4_K_M | 16.39 GB | +15.6 GB | — |
| Mistral Small 24B Instruct24B | Q4_K_M | 15.89 GB | +16.1 GB | 62Estimated |
| Devstral Small 1.1 24B24B | Q6_K | 20.91 GB | +11.1 GB | 48Estimated |
| Magistral Small 1.2 24B24B | Q6_K | 20.91 GB | +11.1 GB | 47Estimated |
| Codestral 22B22B | Q4_K_M | 15.03 GB | +17 GB | 58Estimated |
| ERNIE 4.5 21B-A3B21B-A3B | Q8_0 | 24.44 GB | +7.6 GB | — |
| GPT-OSS 20B21B MoE | MXFP4 | 12.76 GB | +19.2 GB | 195Community |
14B · 16
| Model | Quant | Est. VRAM | Headroom | tok/s on RTX 4090 |
|---|---|---|---|---|
| InternLM2 20B Chat20B | Q5_K_M | 15.59 GB | +16.4 GB | 68Estimated |
| DeepSeek-Coder-V2-Lite Instruct16B-A2.4B | Q8_0 | 17.56 GB | +14.4 GB | 118Estimated |
| DeepSeek-V2-Lite Chat16B-A2.4B | Q4_K_M | 10.08 GB | +21.9 GB | 142Estimated |
| StarCoder2 15B15B | Q4_K_M | 10.19 GB | +21.8 GB | 92Estimated |
| Qwen3 14B Instruct14B | Q5_K_M | 11.65 GB | +20.4 GB | 78 |
| Qwen2.5 14B Instruct14B | Q5_K_M | 11.73 GB | +20.3 GB | 86Estimated |
| DeepSeek-R1-Distill-Qwen-14B14B | Q4_K_M | 10.14 GB | +21.9 GB | 95 |
| Phi-4 14B14B | Q5_K_M | 11.77 GB | +20.2 GB | 78Estimated |
| Phi-3 Medium 14B Instruct14B | Q6_K | 12.86 GB | +19.1 GB | 88Estimated |
| Mistral Nemo 12B Instruct12B | Q6_K | 11.14 GB | +20.9 GB | 95Estimated |
| Gemma 3 12B IT12B | Q5_K_M | 9.84 GB | +22.2 GB | 92Estimated |
| Stable LM 2 12B Chat12B | Q4_K_M | 8.35 GB | +23.7 GB | 108Estimated |
| Llama 3.2 11B Vision Instruct11B | Q8_0 | 12.9 GB | +19.1 GB | 72Estimated |
| Solar 10.7B Instruct11B | Q4_K_M | 7.6 GB | +24.4 GB | 125Estimated |
| Falcon 3 10B Instruct10B | Q4_K_M | 7.28 GB | +24.7 GB | 118Estimated |
| Gemma 2 9B Instruct9B | Q8_0 | 11.7 GB | +20.3 GB | 108Estimated |
7B · 23
| Model | Quant | Est. VRAM | Headroom | tok/s on RTX 4090 |
|---|---|---|---|---|
| GLM-4-9B-Chat9B | Q8_0 | 10.16 GB | +21.8 GB | 105Estimated |
| Qwen3-VL 8B Instruct8B | Q8_0 | 10.39 GB | +21.6 GB | 108Estimated |
| Ministral 3 8B Instruct8B | Q8_0 | 10.02 GB | +22 GB | — |
| Qwen2-VL 7B Instruct7B | Q4_K_M | 5.49 GB | +26.5 GB | 72Estimated |
| Granite 3.1 8B Instruct8B | Q4_K_M | 5.88 GB | +26.1 GB | 142Estimated |
| Qwen3 8B Instruct8B | Q6_K | 7.64 GB | +24.4 GB | 122 |
| Llama 3.1 8B Instruct8B | Q8_0 | 9.47 GB | +22.5 GB | 118 |
| Nous Hermes 3 Llama 3.1 8B8B | Q4_K_M | 5.64 GB | +26.4 GB | 148Estimated |
| Aya 23 8B8B | Q4_K_M | 5.64 GB | +26.4 GB | 145Estimated |
| OpenChat 3.6 8B8B | Q4_K_M | 5.64 GB | +26.4 GB | 146Estimated |
| DeepSeek-R1-Distill-Llama-8B8B | Q5_K_M | 6.51 GB | +25.5 GB | 128Estimated |
| InternLM2 7B Chat7B | Q4_K_M | 5.45 GB | +26.6 GB | 148Estimated |
| Qwen2.5 7B Instruct7B | Q6_K | 6.77 GB | +25.2 GB | 132 |
| Qwen2.5-Coder 7B Instruct7B | Q4_K_M | 5.07 GB | +26.9 GB | 158Estimated |
| DeepSeek-R1-Distill-Qwen-7B7B | Q4_K_M | 5.07 GB | +26.9 GB | 152Estimated |
| Gemma 4 E4B ITE4B | Q8_0 | 8.46 GB | +23.5 GB | — |
| OLMo 2 7B Instruct7B | Q8_0 | 10.3 GB | +21.7 GB | 125Estimated |
| OLMo 3 7B Instruct7B | Q8_0 | 10.3 GB | +21.7 GB | — |
| WizardLM-2 7B7B | Q4_K_M | 5.14 GB | +26.9 GB | 152Estimated |
| Mistral 7B Instruct v0.37B | Q6_K | 6.75 GB | +25.3 GB | 135Estimated |
| Zephyr 7B Beta7B | Q6_K | 6.75 GB | +25.3 GB | 132Estimated |
| Gemma 4 E2B ITE2B | Q8_0 | 5.2 GB | +26.8 GB | — |
| Gemma 3 4B IT4B | Q8_0 | 5.05 GB | +27 GB | 145Estimated |
≤3B · 11
| Model | Quant | Est. VRAM | Headroom | tok/s on RTX 4090 |
|---|---|---|---|---|
| Qwen3 4B Instruct4B | Q6_K | 4.05 GB | +28 GB | 145 |
| Phi-4 Mini Instruct3.8B | Q8_0 | 4.81 GB | +27.2 GB | 262Estimated |
| Phi-3.5 Mini Instruct3.8B | Q8_0 | 5.89 GB | +26.1 GB | 255Estimated |
| Llama 3.2 3B Instruct3B | Q8_0 | 4.05 GB | +28 GB | 285Estimated |
| Qwen2.5 3B Instruct3B | Q8_0 | 3.59 GB | +28.4 GB | 290Estimated |
| SmolLM3 3B3B | Q8_0 | 3.73 GB | +28.3 GB | — |
| Gemma 2 2B Instruct2B | Q8_0 | 3.34 GB | +28.7 GB | 320Estimated |
| Qwen3 1.7B Instruct1.7B | Q8_0 | 2.39 GB | +29.6 GB | 240Estimated |
| Qwen2.5 1.5B Instruct1.5B | Q8_0 | 1.83 GB | +30.2 GB | 410Estimated |
| Llama 3.2 1B Instruct1B | Q8_0 | 1.51 GB | +30.5 GB | 450Estimated |
| Qwen2.5 0.5B Instruct0.5B | Q8_0 | 0.6 GB | +31.4 GB | 540Estimated |
How this list is built
Each row is the lowest-perplexity-loss quant of that model whose estimated total — weights plus KV cache at 4K context plus activation buffer — uses at most 88% of the card. That is the calculator's "green" threshold, so every row here has real headroom rather than only just fitting. Raise the context length and the list shortens; the calculator lets you check any combination directly.
A further 3 models load but with no headroom to spare (up to 105% of VRAM) — the Quant Hub’s GPU chips count those too, which is why its number is higher.
Measured on this card
No benchmark runs in this index were recorded on a Arc Pro B70 32G. Every figure on this page is calculated from the model architecture and the quant level — treat them as estimates, not measurements. The speed column in the tables above is an RTX 4090 figure shown for reference — it is not a speed on this card.
Cards with the same budget
The same 32GB in system memory is a different machine — everything on this page fits there, at a speed set by your DIMMs rather than a graphics bus: 32 GB RAM (CPU)
Setup guides for this card
Step-by-step guides written for the Arc Pro B70 32G's runtime or memory size, with the commands and the checks that prove the model is running on the card.
Common questions
What is the best local LLM for an Arc Pro B70 32G?
For everyday use, GLM-4.7-Flash at Q4_K_M — about 19.2 GB of the card's 32 GB at 4K context, leaving 12.8 GB for a longer window. If you want the largest thing that will load, that is Seed-OSS 36B Instruct at Q5_K_M (27.8 GB, 4.2 GB spare). "Best" here means best fit for the memory budget — this index does not run task benchmarks, so it cannot tell you which model is smarter.
Can an Arc Pro B70 32G run Mixtral 8x7B Instruct?
Only just. Its smallest build here, Q4_K_M, needs about 30.1 GB against 32.0 GB usable — that loads with nothing else running, with no margin for a longer context window. It is not on the list above, which requires a model to stay inside 88% of usable capacity.
How many tokens per second does an Arc Pro B70 32G do on an 8B model at Q4?
This index has no measured run on this card, so it does not publish a figure. What can be stated from specifications: generating a token requires reading every weight once, Llama 3.1 8B Instruct at Q4_K_M is 4.6 GB of weights, and this card moves 608 GB/s — a ceiling near 132 tok/s. Batch size, context length, the runtime and how much of the model sits in cache all take you below it.
Arc Pro B70 32G or RTX 5090 for local LLMs?
They do not hold the same set: Arc Pro B70 32G fits 72 and RTX 5090 74 of 90 at 4K — the cards load different formats. Speed is the other difference — 608 GB/s against 1,792 GB/s, a 2.9× gap in how fast the weights can be read, which is what token generation is bound by. The RTX 5090 generates faster on any model both can hold.
72 of 90 indexed models fit comfortably in 32GB at 4K context, each at the highest-quality quant that still leaves headroom.
This page's figures change when the model or the runtime does.Last updated 2026-10-08 RSS → /feed.xml