Intel · workstation

Arc Pro A30M 4GB

Arc Pro A30M 4GB has 4 GB of VRAM at 112 GB/s — about 3.72 GiB usable after driver and compositor overhead. 842 of 2118 indexed models fit at 8K context with q4_0 KV.

Spec sheet· bandwidth, theoreticalFrom the file· fit from summed bytesPredicted· speed
Memory
4 GB
GDDR6
Bandwidth
112 GB/s
64-bit bus
Tensor FP16
—
dense
TDP
50 W
Context4K8K16K32K64K128K
KV cachef16q8_0q4_0quantizing the KV cache is a ~2× lever on the dominant term at long context
audio asr 37text 699vision language 59video 2audio tts 19embedding 25image 1

What fits at 8K context

largest quantization that fits, per model · 842 of 2118 indexed
ModelBest quantParamsWeights●KV●Total in memory◐Headroom◐tok/s◐
granite-speech-4.1-2b-plusQ4_K2.1B2.75 GiB0.18 GiB3.71 GiB0.01 GiB20±30%
gemma-4-E2B-it-ultra-uncensored-hereticI1-IQ3_S5.1B2.90 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
Barcenas-E2BI1-IQ3_S5.1B2.90 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
gemma-4-E2B-it-Uncensored-MAXI1-IQ3_S5.1B2.90 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
Firefly-v4I1-IQ3_S5.1B2.90 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
gemma-4-E2B-it-abliteratedI1-IQ3_S5.1B2.90 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
rp-model_E2B_v5.1I1-IQ3_S5.1B2.90 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
BartaLens-E2BI1-IQ3_S5.1B2.90 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
EVA-Yi-1.5-9B-32K-V1I1-IQ2_S8.8B2.68 GiB0.21 GiB3.71 GiB0.01 GiB21±30%
gemma-4-E2B-it-uncensoredQ3_K_S5.1B2.90 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
gemma-4-E2B-it-heretic-araQ3_K_S5.1B2.90 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
gemma-4-E2BQ3_K_S5.1B2.90 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
Fara1.5-4BQ4_K_L4.5B2.83 GiB0.07 GiB3.71 GiB0.01 GiB21±30%
AREX-TurboQ4_K_L4.5B2.83 GiB0.07 GiB3.71 GiB0.01 GiB21±30%
gemma-3n-E2B-itQ4_15.4B2.87 GiB0.04 GiB3.71 GiB0.01 GiB21±30%
gemma-4-E2B-it-qat-q4_0-unquantized-hereticI1-IQ3_S5.1B2.89 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
Huihui-gemma-4-E2B-it-qat-q4_0-unquantized-abliteratedI1-IQ3_S5.1B2.89 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
Gemma4_E2B_Abliterated_Baked_HF_ReadyI1-IQ3_S5.1B2.89 GiB0.02 GiB3.71 GiB0.01 GiB21±30%
gemma-4-E2B-itIQ3_XS5.1B2.89 GiB0.02 GiB3.70 GiB0.02 GiB21±30%
Darwin-4B-ChimeraI1-Q5_K_M4.0B2.69 GiB0.20 GiB3.70 GiB0.02 GiB21±30%
Voxtral-Mini-3B-2507Q5_K_S4.7B2.63 GiB0.26 GiB3.70 GiB0.02 GiB21±30%
granite-speech-4.1-2bQ4_K2.3B2.74 GiB0.18 GiB3.70 GiB0.02 GiB21±30%
umt5-xxlQ3_K_M5.7B2.85 GiB0.00 GiB3.70 GiB0.02 GiB21±30%
Ministral-3-8B-Instruct-2512-BF16-abliteratedI1-IQ2_XS8.9B2.56 GiB0.30 GiB3.70 GiB0.02 GiB21±30%
Amaretto-8BI1-IQ2_XS8.9B2.56 GiB0.30 GiB3.70 GiB0.02 GiB21±30%
Llama-3.2-3B-Instruct-uncensoredQ5_K_L3.6B2.64 GiB0.25 GiB3.70 GiB0.02 GiB21±30%
Miril-Drone-2B-1IQ3_XS5.1B2.88 GiB0.02 GiB3.70 GiB0.02 GiB21±30%
MARTHA-LXVII.8B_QWEN-3.5-3.6_prune_9b-3.6_baseI1-IQ2_XS8.1B2.80 GiB0.06 GiB3.69 GiB0.03 GiB21±30%
medgemma-4b-itQ5_K_L4.3B2.79 GiB0.09 GiB3.69 GiB0.03 GiB21±30%
amoral-gemma3-4B-v1Q5_K_L4.3B2.79 GiB0.09 GiB3.69 GiB0.03 GiB21±30%
gemma-3-4b-it-abliteratedQ5_K_L4.3B2.79 GiB0.09 GiB3.69 GiB0.03 GiB21±30%
gemma-3-4b-itQ5_K_L4.3B2.79 GiB0.09 GiB3.69 GiB0.03 GiB21±30%
EXAONE-4.0-1.2B-abliteratedF161.5B2.78 GiB0.13 GiB3.69 GiB0.03 GiB21±30%
whisper-mediumF32764M2.85 GiB0.00 GiB3.69 GiB0.03 GiB21±30%
whisper-medium.enF32764M2.85 GiB0.00 GiB3.69 GiB0.03 GiB21±30%
INTELLECT-1-InstructI1-IQ1_M10.2B2.48 GiB0.37 GiB3.69 GiB0.03 GiB21±30%
Phi-3.5-mini-instructQ4_K_S3.8B2.04 GiB0.84 GiB3.69 GiB0.03 GiB21±30%
Foundation-Sec-8B-InstructI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Foundation-Sec-8B-Instruct-hereticI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Meta-Llama-3-8BIQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
llama3.1-heretic-unsensoredI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Anubis-Mini-8B-v1I1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
dolphin-2.9-llama3-8bIQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Meta-Llama-3-8BIQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
L3.1-Dark-Reasoning-LewdPlay-evo-Hermes-R1-Uncensored-8B-hereticI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
L3.1-Dark-Reasoning-LewdPlay-evo-Hermes-R1-Uncensored-8BI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Gluon-8BI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Llama3.3-8B-Instruct-Thinking-Heretic-Uncensored-Claude-4.5-Opus-High-ReasoningI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Llama3.3-8B-Instruct-Thinking-Claude-4.5-Opus-High-ReasoningI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Hypnos-i1-8BI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
grok-oss-Apollyon-8B-hereticI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Llama-3.3-8B-Instruct-128K-JbliteratedI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
PsyCoPref-Llama3-8BI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Llama3.1-GptDeluxe-8BI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
SelfCite-8B-CC-SFTI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
ai-girlfriend-v2I1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Llama-3.3-8B-Instruct-128K_AbliteratedI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
Aisha-Uncensored-v2I1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
grok-oss-Apollyon-8BI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
HARC-Llama-3.1-8B-InstructI1-IQ2_S8.0B2.57 GiB0.28 GiB3.69 GiB0.03 GiB21±30%
●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 Pro A30M 4GB run?
842 of 2118 indexed open-weight models fit a Arc Pro A30M 4GB at 8,192 context with q4_0 KV cache, the largest being granite-speech-4.1-2b-plus at Q4_K. That covers text, vision-language, image, video and speech models.
How much usable memory does a Arc Pro A30M 4GB actually have?
Its nameplate is 4 GB, but about 3.72 GiB is available to a model once driver and compositor overhead is accounted for.
Is a Arc Pro A30M 4GB fast for local AI?
Its memory bandwidth is 112 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.