RunsOnMy ModelsGPUs & MacsAPI pricesVRAM calculatorNew

What AI models run on the RTX A6000?

The RTX A6000 has 48 GB of VRAM (768 GB/s). About 45.6 GB is usable for a model. 214 of 285 open models we track fit at 8K context.

214models fit (8K ctx)
212still fit at 32K ctx
127Blargest that fits
768 GB/smemory bandwidth

Models that fit, largest first

ModelParamsBest quant (8K)Needs32K contextMax tok/s (est.)
Ling 3.0 flash127B (5.6B act.)Q2_K43.8 GBQ2_K381details
Ornith 1.5 122B125B (13B act.)Q2_K44.7 GBQ2_K161details
GLM 4.5 Air110B (15B act.)Q2_K44.3 GBno125details
Llama 4 Scout 17B 16E Instruct109B (18B act.)Q2_K39.1 GBQ2_K117details
Llama 4 Scout 17B 16E109B (18B act.)Q2_K40.2 GBQ2_K113details
GLM 4.5V108B (14B act.)Q2_K39.7 GBQ2_K143details
GLM 4.6V108B (14B act.)Q2_K43.0 GBno132details
Hunyuan A13B Instruct80B (13B act.)Q3_K_M37.6 GBQ3_K_M123details
Kimi Dev 72B73BQ3_K_M38.3 GBQ2_K20details
Llama 3.3 70B Instruct71BQ3_K_M35.3 GBQ2_K22details
DeepSeek R1 Distill Llama 70B71BQ3_K_M35.3 GBQ2_K22details
Llama 3.1 70B71BQ3_K_M35.1 GBQ2_K22details
Kimi Linear 48B A3B Instruct49B (3.0B act.)Q6_K38.6 GBQ6_K311details
Nous Hermes 2 Mixtral 8x7B DPO47B (13B act.)Q6_K37.2 GBQ6_K73details
falcon 40b42BQ6_K33.6 GBQ6_K22details
flint40B (1.3B act.)Q8_040.3 GBQ8_0578details
Maxwell V1 Flash37B (2.3B act.)Q8_037.6 GBQ8_0318details
Hermes 4.3 36B36BQ8_038.5 GBQ6_K20details
Seed OSS 36B Instruct36BQ8_038.5 GBQ6_K20details
Qwen3.6 35B A3B36B (3.0B act.)Q8_036.2 GBQ8_0243details
Ornith 1.5 35B A3B36B (3.0B act.)Q8_036.2 GBQ8_0243details
Qwen3.5 35B A3B36B (3.0B act.)Q8_035.4 GBQ8_0249details
Axion 1.0 35B35B (3.9B act.)Q8_035.8 GBQ8_0185details
Qwen AgentWorld 35B A3B35B (3.0B act.)Q8_035.4 GBQ8_0240details
Yi 34B 200K34BQ8_036.4 GBQ6_K21details
Yi 34B34BQ8_036.4 GBQ6_K21details
Yi 34B Chat34BQ8_036.4 GBQ6_K21details
Yi 1.5 34B Chat34BQ8_036.4 GBQ6_K21details
Nous Hermes 2 Yi 34B34BQ8_036.4 GBQ6_K21details
DeepSeek R1 Distill Qwen 32B33BQ8_035.1 GBQ6_K22details
Qwen3 32B33BQ8_035.1 GBQ6_K22details
GLM Z1 32B 041433BQ8_033.4 GBQ8_022details
GLM 4 32B 041433BQ8_033.4 GBQ8_022details
granite 4.0 h small32B (8.8B act.)Q8_032.6 GBQ8_082details
gemma 4 31B it31BQ8_033.3 GBQ8_024details
GLM 4.7 Flash31B (4.6B act.)Q8_030.8 GBQ8_0163details
Qwen3 Coder 30B A3B Instruct31B (3.0B act.)Q8_031.7 GBQ8_0241details
North Mini Code 1.030B (3.3B act.)Q8_031.6 GBQ8_0218details
Hy Memory 30B A3B30B (3.0B act.)Q8_031.1 GBQ8_0241details
Muse Glimmer 30B30BQ8_028.5 GBQ8_026details
granite 4.2 30b29BQ8_031.6 GBQ8_025details
Qwen3.8 27B28BQ8_028.4 GBQ8_026details
Qwen3.6 27B28BQ8_028.0 GBQ8_027details
Qwen3.5 27B28BQ8_028.0 GBQ8_027details
gemma 3 27b it27BQ8_028.7 GBQ8_027details
Arete 27B27BQ8_028.1 GBQ8_027details
MiniMax H3 Image PE27BQ8_028.0 GBQ8_027details
Trinity Mini26B (2.0B act.)Q8_026.9 GBQ8_0367details
GLiDE 27B26BQ8_027.2 GBQ8_028details
diffusiongemma 26B A4B it26B (4.0B act.)Q8_026.4 GBQ8_0184details
gemma 4 26B A4B it26B (4.0B act.)Q8_026.4 GBQ8_0184details
LFM2.5 24B A2B24B (2.0B act.)Q8_024.5 GBQ8_0361details
ERNIE 4.5 21B A3B PT22BQ8_022.6 GBQ8_033details
ERNIE 4.5 21B A3B Thinking22BQ8_022.6 GBQ8_033details
gpt oss safeguard 20b22B (4.8B act.)F1613.8 GBF16250details
gpt oss 20b21B (4.2B act.)F1613.8 GBF16278details
Groot 19B A3B 26102319B (3.0B act.)F1636.4 GBF16131details
Kimi VL A3B Thinking 250616B (3.0B act.)Q8_016.8 GBQ8_0247details
Kimi VL A3B Instruct16B (3.0B act.)F1631.5 GBF16128details
Kimi VL A3B Thinking16B (3.0B act.)F1631.5 GBF16128details
deepseek moe 16b base16B (2.8B act.)F1632.8 GBF16136details
LLaDA2.1 mini16B (1.4B act.)F1631.3 GBF16268details
Moonlight 16B A3B Instruct16B (3.0B act.)F1630.7 GBF16128details
DeepSeek Coder V2 Lite Instruct16B (2.7B act.)F1630.1 GBF16144details
DeepSeek V2 Lite16B (2.7B act.)F1630.1 GBF16144details
DeepSeek V2 Lite Chat16B (2.7B act.)F1630.1 GBF16144details
Qwen2.5 Coder 14B Instruct15BF1629.7 GBF1626details
DeepSeek R1 Distill Qwen 14B15BF1629.7 GBF1626details
Qwen3 14B15BF1629.4 GBF1626details
phi 415BF1629.5 GBF1626details
gemma 3 12b it12BF1623.6 GBF1633details
Llama Guard 4 12B12BF1624.6 GBF1632details
Llama 3.2 11B Vision Instruct11BF1620.1 GBF1639details
Llama 3.2 11B Vision11BF1621.7 GBF1636details
GLM 4.6V Flash10BF1618.5 GBF1641details
GLM 4.1V 9B Thinking10BF1618.5 GBF1641details
Step3 VL 10B10BF1632.3 GBF1623details
SenseNova Loop 9B Preview9.7BF1617.8 GBF1643details
Qwen3.5 9B9.7BF1617.8 GBF1643details
Ornith 1.5 9B9.7BF1617.8 GBF1643details
Qwen3.5 9B Base9.7BF1617.8 GBF1643details
MiMo V2.6 Distill Qwen 9B9.4BF1617.8 GBF1643details
glm 4 9b9.4BF1623.2 GBF1641details
internlm3 8b instruct8.8BF1617.4 GBF1644details
granite 4.2 8b8.8BF1618.2 GBF1644details
granite 4.1 8b8.8BF1618.2 GBF1644details
Qwen3 VL 8B Instruct8.8BF1617.1 GBF1647details
Molmo2 8B8.7BF1617.1 GBF1647details
LFM2.5 8B A1B8.5B (1.0B act.)F1632.3 GBF16192details
Qwen2.5 VL 7B Instruct8.3BF1615.3 GBF1650details
UI TARS 1.5 7B8.3BF1615.3 GBF1650details
olmOCR 2 7B 10258.3BF1615.3 GBF1650details
olmOCR 7B 0225 preview8.3BF1616.6 GBF1646details
UI TARS 7B DPO8.3BF1615.3 GBF1650details
Qwen3 8B8.2BF1617.1 GBF1647details
DeepSeek R1 0528 Qwen3 8B8.2BF1617.1 GBF1647details
granite 3.3 8b instruct8.2BF1617.0 GBF1647details
granite 3.1 8b instruct8.2BF1617.0 GBF1647details
granite 3.2 8b instruct8.2BF1617.0 GBF1647details
Apertus 8B Instruct 25098.1BF1616.7 GBF1648details
Llama 3.1 8B Instruct8.0BF1616.6 GBF1648details
Llama 3.1 8B8.0BF1616.6 GBF1648details
Hermes 3 Llama 3.1 8B8.0BF1616.6 GBF1648details
DeepSeek R1 Distill Llama 8B8.0BF1631.6 GBF1624details
Meta Llama 3 8B Instruct8.0BF1616.6 GBF1648details
Meta Llama Guard 2 8B8.0BF1616.6 GBF1648details
Meta Llama 3.1 8B Instruct8.0BF1616.6 GBF1648details
Ling 3.0 tiny7.9B (1.4B act.)F1615.6 GBF16278details
MiMo 7B RL7.8BF1616.4 GBF1649details
internlm2 5 7b chat7.7BF1616.0 GBF1650details
internlm2 chat 7b7.7BF1616.0 GBF1650details
Qwen2.5 7B Instruct7.6BF1615.3 GBF1650details
Qwen2.5 Coder 7B Instruct7.6BF1615.3 GBF1650details
DeepSeek R1 Distill Qwen 7B7.6BF1615.3 GBF1650details
Falcon H1R 7B7.6BF1615.1 GBF1651details
Olmo 3 7B Instruct7.3BF1616.7 GBF1653details
Olmo 3 1025 7B7.3BF1616.7 GBF1653details
Olmo 3 7B Think7.3BF1616.7 GBF1653details
Mistral 7B v0.17.2BF1614.9 GBF1653details
Mistral 7B Instruct v0.27.2BF1614.9 GBF1653details
Mistral 7B Instruct v0.17.2BF1614.9 GBF1653details
falcon 7b7.2BF1618.4 GBF1653details
falcon 7b instruct7.2BF1618.4 GBF1653details
granite 4.0 h tiny6.9B (1.5B act.)F1613.6 GBF16262details
OLMoE 1B 7B 09246.9B (1.3B act.)F1614.4 GBF16299details
deepseek coder 7b instruct v1.56.9BF1617.2 GBF1656details
deepseek coder 6.7b instruct6.7BF1617.0 GBF1657details
deepseek coder 6.7b base6.7BF1617.0 GBF1657details
Llama 2 7b hf6.7BF1617.0 GBF1657details
Yi 6B 200K6.1BF1612.3 GBF1663details
Yi 6B6.1BF1612.3 GBF1663details
Molmo2 4B4.9BF1610.0 GBF1687details
WeVisOne VL 4B4.8BF1610.0 GBF1687details
Mage VL Cowork4.7BF1610.0 GBF1687details
Qianfan OCR4.7BF1610.1 GBF1687details
Qwen3.5 4B4.7BF169.0 GBF1691details
S5 nano4.7BF169.2 GBF1689details
LightOnOCR 3 4B4.5BF169.0 GBF1691details
SingVL S1 4B4.5BF169.0 GBF1691details
Intern Decision 4B4.5BF169.0 GBF1691details
Qwen3 VL 4B Instruct4.4BF169.3 GBF1695details
gemma 3 4b it4.3BF168.4 GBF1699details
medgemma 1.5 4b it4.3BF168.4 GBF1699details
medgemma 4b it4.3BF168.4 GBF1699details
gemma 3 4b pt4.3BF169.2 GBF1689details
Phi 3.5 vision instruct4.1BF1611.2 GBF1693details
Qwen3 4B4.0BF169.3 GBF1695details
Qwen3 4B Instruct 25074.0BF169.3 GBF1695details
granite vision 4.1 4b4.0BF167.6 GBF16113details
Phi 4 mini instruct3.8BF168.9 GBF16100details
Phi 3 mini 4k instruct3.8BF1610.6 GBF16100details
Phi 3.5 mini instruct3.8BF1610.6 GBF16100details
Phi 3 mini 128k instruct3.8BF1610.6 GBF16100details
Qwen2.5 VL 3B Instruct3.8BF166.7 GBF16124details
granite 4.2 3b3.7BF168.0 GBF16105details
granite 4.1 3b llmon preview3.7BF168.0 GBF16105details
Raon 1 nano3.5BF167.0 GBF16122details
granite 4.1 3b3.4BF167.6 GBF16113details
granite 4.0 micro3.4BF167.6 GBF16113details
DeepSeek OCR 23.4B (1.2B act.)F166.6 GBF16371details
deepseek vl2 tiny3.4B (1.2B act.)F167.4 GBF16327details
Unlimited OCR 23.3B (1.1B act.)F166.6 GBF16380details
Unlimited OCR3.3B (1.1B act.)F166.6 GBF16382details
DeepSeek OCR3.3B (1.1B act.)F166.6 GBF16382details
granite 3.1 3b a800m instruct3.3B (883M act.)F167.1 GBF16435details
Llama 3.2 3B Instruct3.2BF1613.5 GBF1660details
Llama 3.2 3B3.2BF167.5 GBF16120details
Hermes 3 Llama 3.2 3B3.2BF167.5 GBF16119details
d1 3B3.1BF1610.8 GBF1671details
LFM2.5 VL 3B3.1BF1610.8 GBF1671details
Qwen2.5 3B Instruct3.1BF166.7 GBF16124details
SmolLM3 3B3.1BF166.9 GBF16125details
SmolLM3 3B Base3.1BF166.9 GBF16125details
phi 22.8BF168.2 GBF16138details
LFM2.5 2.6B2.7BF1610.8 GBF1671details
granite 3.3 2b instruct2.5BF165.8 GBF16152details
MiniCPM5 2B2.5BF165.7 GBF16152details
WeVisOne VL 2B2.4BF165.4 GBF16188details
Qwen3.5 2B2.3BF164.5 GBF16203details
SmolVLM Instruct2.2BF165.4 GBF16212details
Qwen3 1.7B2.0BF164.8 GBF16223details
Qwen3.8 27B DSpine b162.0BF164.8 GBF16189details
d1 2B1.9BF167.2 GBF16111details
LFM3 VL 2B1.9BF167.2 GBF16111details
DeepSeek R1 Distill Qwen 1.5B1.8BF164.2 GBF16216details
LFM3 1.7B1.7BF167.2 GBF16111details
LFM3 1.7B Base1.7BF164.0 GBF16222details
SmolLM2 1.7B Instruct1.7BF165.2 GBF16224details
SmolLM2 1.7B1.7BF165.2 GBF16224details
SmolLM 1.7B1.7BF165.2 GBF16224details
granite 4.0 1b1.6BF164.3 GBF16235details
Qwen2.5 1.5B Instruct1.5BF163.8 GBF16249details
OLMo 2 0425 1B1.5BF164.4 GBF16258details
deepseek coder 1.3b instruct1.3BF164.5 GBF16285details
Llama 3.2 1B1.2BF163.2 GBF16311details
Llama 3.2 1B Instruct1.2BF165.5 GBF16155details
LFM2.5 1.2B Instruct1.2BF165.0 GBF16164details
HunyuanOCR1.1BF163.5 GBF16343details
PaddleOCR VL 1.6959MF162.5 GBF16401details
Qwen3.5 0.8B873MF162.4 GBF16506details
LightOnOCR 3 0.8B853MF162.4 GBF16506details
FastVLM 0.5B759MF162.2 GBF16506details
Qwen3 0.6B752MF162.7 GBF16641details
GOT OCR2 0716MF162.8 GBF16536details
Nemotron 3.5 Parse631MF161.9 GBF16588details
GOT OCR 2.0 hf561MF162.5 GBF16685details
SmolVLM2 500M Video Instruct507MF161.6 GBF16936details
SmolVLM 500M Instruct507MF161.6 GBF16936details
Qwen2.5 0.5B Instruct494MF161.7 GBF16777details
LFM2.5 VL 450M449MF161.9 GBF16540details
SmolLM2 360M Instruct362MF161.5 GBF161059details
SmolLM2 360M362MF161.5 GBF161059details
LFM2.5 350M354MF161.9 GBF16540details
LFM2 350M354MF161.4 GBF161083details

Just too big at 4-bit

These need up to 35% more memory than the RTX A6000 has at Q4_K_M and 8K context. A smaller quant or shorter context may work.

ModelQ4_K_M needsShort by
Hunyuan A13B Instruct47.1 GB1.5 GB
Mistral 7B Instruct v0.247.7 GB2.1 GB
gpt oss safeguard 120b59.5 GB13.9 GB
gpt oss 120b59.5 GB13.9 GB
GLM 4.5V61.4 GB15.8 GB

Questions

What is the biggest model that runs on the RTX A6000?

Of the models we track, the largest that fits at 8K context is Ling 3.0 flash (127B) at Q2_K, needing about 43.8 GB.

How much of the RTX A6000's memory can a model use?

About 45.6 GB: we assume 95% of the 48 GB is usable after driver and display overhead.

Spec source: NVIDIA specifications. Model data: Hugging Face Hub API. Speed is a bandwidth ceiling (bandwidth ÷ bytes read per token), not a benchmark. Methodology · Report an error.