Intel · consumer

Arc B570 10GB

Arc B570 10GB has 10 GB of VRAM at 380 GB/s — about 9.30 GiB usable after driver and compositor overhead. 1187 of 2118 indexed models fit at 128K context with q4_0 KV.

Spec sheet· bandwidth, theoreticalFrom the file· fit from summed bytesPredicted· speed
Memory
10 GB
GDDR6
Bandwidth
380 GB/s
160-bit bus
Tensor FP16
—
dense
TDP
150 W
$219 MSRP
Context4K8K16K32K64K128K
KV cachef16q8_0q4_0quantizing the KV cache is a ~2× lever on the dominant term at long context
text 990vision language 100embedding 26image 1video 12audio asr 38audio tts 20

What fits at 128K context

largest quantization that fits, per model · 1187 of 2118 indexed
ModelBest quantParamsWeights●KV●Total in memory◐Headroom◐tok/s◐
Gemma-4-12B-StyleTuneI1-Q3_K_M13.0B6.07 GiB2.38 GiB9.30 GiB0.00 GiB24±30%
gemma-4-12b-heretic-styletune-headI1-Q3_K_M12.0B6.07 GiB2.38 GiB9.30 GiB0.00 GiB24±30%
syrian-gemma-12bI1-Q3_K_M13.0B6.07 GiB2.38 GiB9.30 GiB0.00 GiB24±30%
OmniAtlas-Qwen3-30B-A3BI1-IQ2_XS31.7B8.45 GiB0.00 GiB9.30 GiB0.00 GiB24±30%
Qwen3-Omni-30B-A3B-CaptionerI1-IQ2_XS31.7B8.45 GiB0.00 GiB9.30 GiB0.00 GiB24±30%
LFM2-24B-A2BMoEQ2_K_L23.8B7.78 GiB0.70 GiB9.30 GiB0.00 GiB65±37%
Trinity-MiniMoEIQ2_M26.1B7.90 GiB0.60 GiB9.29 GiB0.01 GiB68±37%
MiMo-VL-7B-RLI1-IQ3_M8.3B3.40 GiB5.06 GiB9.29 GiB0.01 GiB24±30%
Kuwutu-7B-CYOA-v2I1-IQ3_M7.6B3.40 GiB5.06 GiB9.29 GiB0.01 GiB24±30%
Marco-Nano-InstructMoEI1-Q4_K_S8.0B4.57 GiB3.94 GiB9.28 GiB0.02 GiB26±37%
Aya-Medikal-V2I1-Q3_K_M8.0B3.93 GiB4.50 GiB9.28 GiB0.02 GiB24±30%
granite-3.1-8b-instructQ2_K_S8.2B2.82 GiB5.63 GiB9.28 GiB0.02 GiB24±30%
Skywork-R1V3-38BIQ2_XXS38.4B8.41 GiB0.00 GiB9.27 GiB0.03 GiB24±30%
qwen-indic-v1I1-IQ3_M7.6B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
next-8bI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Supertron2-Reranker-8BI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
next-ocrI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen3-VL-8B-GLM-4.7-Flash-Heretic-Uncensored-ThinkingI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Midas-FableAgent-8BI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen3-VL-8B-Heretic-1.3.0I1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen3-VL-8B-Thinking-Unredacted-MAXI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen3-VL-8B-Instruct-Minecraft-MT-en-zhI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen-3-VL-8B-Instruct-hereticI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Poe-8B-GLM5-Opus4.6-Sonnet4.5-Kimi-Grok-Gemini-3-pro-preview-HERETICI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
ToolCUA-8BI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Huihui-Qwen3-VL-8B-Instruct-abliteratedI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen3-VL-Reranker-8BI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Salience-1-9BI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen3-VL-8B-Instruct-Uncensored-V2I1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Maestro1-9BI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
GRaPE-2-FlashI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Jan-v2-VL-medI1-IQ3_XS8.8B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Parable-Qwen3-8B-Claude-Fable-5I1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
ReasonCritic-7BI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
mythos-9b-unhinged-hereticI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Finch-8B-KTOI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Finch-8BI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
MathSmith-hc-Qwen3-8BI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
MiroThinker-v1.0-8BI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
mythos-9b-unhingedI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Ektome-Qwen3-8B-PristinelyUncensoredI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Marco-DeepResearch-8BI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
mythos-9b-mergedI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
qwen3-8b-apostateI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Josiefied-Qwen3-8B-abliterated-v1I1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
tmax-8bI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen3-8B-abliteratedIQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen3-8BIQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen3-8B-DeepSeek-v3.2-Speciale-DistillIQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
story_generation_Qwen3_8B_RLI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
AReaL-boba-2-8B-OpenI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
DS-R1-Qwen3-8B-ArliAI-RpR-v4-SmallI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Nemotron-Orchestrator-8BIQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
DeepSeek-R1-0528-Qwen3-8BIQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
S1-Base-8BI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Huihui-Qwen3-8B-abliterated-v2I1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Step3-VL-10B-BaseI1-IQ3_XS10.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
Qwen3-Reranker-8BI1-IQ3_XS8.2B3.38 GiB5.06 GiB9.27 GiB0.03 GiB24±30%
HunyuanImage-2.1Q3_K_S17.5B8.42 GiB0.00 GiB9.27 GiB0.03 GiB24±30%
OLMoE-1B-7B-0924-InstructMoEQ4_K_L6.9B4.00 GiB4.50 GiB9.27 GiB0.03 GiB22±37%
●From the file◐Predictedwhat these mean

Speed is modeled, not measured: decode is memory-bandwidth bound, so tokens per second is bytes read per token against achievable bandwidth. Mixture-of-experts models carry a wider band because only the routed experts are read each step, and few have been measured publicly.

Questions people ask

What AI models can a Arc B570 10GB run?
1187 of 2118 indexed open-weight models fit a Arc B570 10GB at 131,072 context with q4_0 KV cache, the largest being Gemma-4-12B-StyleTune at I1-Q3_K_M. That covers text, vision-language, image, video and speech models.
How much usable memory does a Arc B570 10GB actually have?
Its nameplate is 10 GB, but about 9.30 GiB is available to a model once driver and compositor overhead is accounted for.
Is a Arc B570 10GB fast for local AI?
Its memory bandwidth is 380 GB/s, and that figure — not teraflops — is what governs token generation speed. Capacity decides what you can run; bandwidth decides how fast it runs.