RunsOnMy ModelsGPUs & MacsAPI pricesVRAM calculatorNew

VRAM requirements for 285 open AI models

Memory needed at 8K context, weights + KV cache + overhead. Sorted by downloads. Type to filter.

ModelOrgParamsQ4_K_MQ8_0AttentionMax ctxOn HF
Qwen3 0.6BQwen752M1.9 GB2.2 GBGQA40KApr 27, 2025
Qwen3 14BQwen15B10.3 GB16.6 GBGQA40KApr 27, 2025
DeepSeek V3.2deepseek-ai685B MoE379 GB666 GBMLA160KDec 1, 2025
Qwen3 VL 8B InstructQwen8.8B6.5 GB9.9 GBGQA256KOct 11, 2025
Qwen3.8 27BQwen28B16.8 GB28.4 GBGQA256KAug 5, 2026
Qwen2.5 7B InstructQwen7.6B5.5 GB8.7 GBGQA32KSep 16, 2024
GLM 5.3 Flashzai-org321B MoE184 GB318 GBMLA1MAug 25, 2026
gemma 4 31B itgoogle31B20.0 GB33.3 GBGQA + sliding window256KMar 11, 2026
Qwen3.5 9BQwen9.7B6.4 GB10.0 GBGQA256KFeb 27, 2026
Qwen3 8BQwen8.2B6.5 GB9.9 GBGQA40KApr 27, 2025
Unlimited OCRbaidu3.3B MoE2.9 GB4.0 GBMHA32KJun 19, 2026
gpt oss 20bopenai21B MoE11.8 GB12.3 GBGQA + sliding window128KAug 4, 2025
Qwen2.5 VL 7B InstructQwen8.3B5.5 GB8.7 GBGQA125KJan 26, 2025
gemma 4 26B A4B itgoogle26B MoE17.2 GB26.4 GBGQA + sliding window256KMar 11, 2026
Qwen2.5 1.5B InstructQwen1.5B1.9 GB2.7 GBGQA32KSep 17, 2024
Qwen3.5 2BQwen2.3B2.2 GB2.8 GBGQA256KFeb 28, 2026
Qwen3.5 4BQwen4.7B3.7 GB5.3 GBGQA256KFeb 27, 2026
Llama 3.1 8B Instructmeta-llama8.0B6.2 GB9.6 GBGQA128KJul 18, 2024
Qwen2.5 0.5B InstructQwen494M1.2 GB1.4 GBGQA32KSep 16, 2024
Qwen3.6 35B A3BQwen36B MoE22.1 GB36.2 GBGQA256KApr 15, 2026
Qwen3 32BQwen33B21.1 GB35.1 GBGQA40KApr 27, 2025
Qwen3 VL 4B InstructQwen4.4B4.1 GB5.8 GBGQA256KOct 11, 2025
Qwen3 4BQwen4.0B4.1 GB5.8 GBGQA40KApr 27, 2025
Qwen3 1.7BQwen2.0B2.6 GB3.3 GBGQA40KApr 27, 2025
Qwen3.8 Flash NextQwen180B MoE104 GB176 GBGQA256KAug 24, 2026
Qwen3.5 0.8BQwen873M1.5 GB1.7 GBGQA256KFeb 28, 2026
Qwen3 4B Instruct 2507Qwen4.0B4.1 GB5.8 GBGQA256KAug 5, 2025
Qwen3.6 27BQwen28B17.0 GB28.0 GBGQA256KApr 21, 2026
gpt oss 120bopenai117B MoE59.5 GB60.1 GBGQA + sliding window128KAug 4, 2025
DeepSeek V4 Flash 0731deepseek-ai304B MoE175 GB302 GBGQA1MJul 31, 2026
Qwen2.5 3B InstructQwen3.1B2.9 GB4.3 GBGQA32KSep 17, 2024
Qwen3 Coder 30B A3B InstructQwen31B MoE18.7 GB31.7 GBGQA256KJul 31, 2025
Qwen2.5 VL 3B InstructQwen3.8B2.8 GB4.0 GBGQA125KJan 26, 2025
DeepSeek OCRdeepseek-ai3.3B MoE2.9 GB4.0 GBMHA8KOct 17, 2025
GLM 5.3zai-org753B MoE430 GB748 GBMLA1MAug 25, 2026
Qwen2.5 Coder 14B InstructQwen15B10.6 GB16.8 GBGQA32KNov 6, 2024
SmolVLM2 500M Video InstructHuggingFaceTB507M1.1 GB1.2 GBGQA8KFeb 11, 2025
Qwen3.5 27BQwen28B17.0 GB28.0 GBGQA256KFeb 24, 2026
DeepSeek V4.1 Flashdeepseek-ai763B MoE436 GB757 GBGQA1MSep 10, 2026
DeepSeek R1deepseek-ai684B MoE378 GB665 GBMLA160KJan 20, 2025
Qwen2.5 Coder 7B InstructQwen7.6B5.5 GB8.7 GBGQA32KSep 17, 2024
Inklingthinkingmachines952B MoE545 GB801 GBGQA—Jul 14, 2026
Qwen3.5 35B A3BQwen36B MoE21.5 GB35.4 GBGQA256KFeb 24, 2026
DeepSeek V3deepseek-ai684B MoE378 GB665 GBMLA160KDec 25, 2024
Llama 3.2 1B Instructmeta-llama1.2B1.6 GB2.1 GBGQA128KSep 18, 2024
DeepSeek V4 Flashdeepseek-ai291B MoE167 GB289 GBGQA1MApr 22, 2026
GLM 5zai-org754B MoE426 GB748 GBMLA198KFeb 11, 2026
Mistral 7B Instruct v0.2mistralai7.2B47.7 GB8.6 GBGQA32KDec 11, 2023
GLM 4.7 Flashzai-org31B MoE18.2 GB30.8 GBMLA198KJan 19, 2026
GLM 4.5 Airzai-org110B MoE70.1 GB112 GBGQA128KJul 20, 2025
MiniMax M2.5MiniMaxAI229B MoE132 GB229 GBGQA192KFeb 12, 2026
MiniCPM5 2Bopenbmb2.5B2.5 GB3.5 GBGQA128KSep 6, 2026
gemma 3 4b itgoogle4.3B3.5 GB5.0 GBGQA + sliding window128KFeb 20, 2025
Hy3tencent299B MoE173 GB299 GBGQA256KJul 2, 2026
Llama 3.2 3B Instructmeta-llama3.2B3.4 GB4.7 GBGQA128KSep 18, 2024
Kimi K3moonshotai2.78T MoE1584 GB2752 GBMLA1MJun 13, 2026
GLM 5.1zai-org754B MoE434 GB748 GBMLA198KApr 3, 2026
diffusiongemma 26B A4B itgoogle26B MoE17.1 GB26.4 GBGQA + sliding window256KJun 9, 2026
HunyuanOCRtencent1.1B2.0 GB2.5 GBGQA128KNov 18, 2025
Llama 3.2 1Bmeta-llama1.2B1.8 GB2.4 GBGQA128KSep 18, 2024
Ornith 1.5 9Bornith-ai9.7B6.6 GB10.0 GBGQA256KAug 18, 2026
Unlimited OCR 2baidu3.3B MoE3.0 GB4.0 GBMHA128KOct 26, 2026
DeepSeek R1 0528 Qwen3 8Bdeepseek-ai8.2B6.5 GB9.9 GBGQA128KMay 29, 2025
DeepSeek V4.1 Flash 1017deepseek-ai763B MoE436 GB757 GBGQA1MOct 17, 2026
medgemma 4b itgoogle4.3B3.5 GB5.0 GBGQA + sliding window128KMay 19, 2025
Mistral 7B v0.1mistralai7.2B5.5 GB8.6 GBGQA32KSep 20, 2023
Llama 3.1 8Bmeta-llama8.0B6.2 GB9.6 GBGQA128KJul 14, 2024
DeepSeek R1 Distill Qwen 1.5Bdeepseek-ai1.8B1.9 GB2.7 GBGQA128KJan 20, 2025
granite 4.1 3bibm-granite3.4B3.2 GB4.6 GBGQA128KApr 6, 2026
gemma 3 12b itgoogle12B8.5 GB13.4 GBGQA + sliding window128KMar 1, 2025
Kimi K2.5moonshotai1.03T MoE580 GB1017 GBMLA256KJan 1, 2026
DeepSeek R1 Distill Qwen 32Bdeepseek-ai33B21.2 GB35.1 GBGQA128KJan 20, 2025
Ornith 1.5 122Bornith-ai125B MoE72.3 GB125 GBGQA256KOct 16, 2026
Llama 3.3 70B Instructmeta-llama71B43.3 GB73.0 GBGQA128KNov 26, 2024
Hermes 3 Llama 3.1 8BNousResearch8.0B6.2 GB9.6 GBGQA128KJul 28, 2024
Phi 3.5 mini instructmicrosoft3.8B5.7 GB7.2 GBMHA128KAug 16, 2024
Ornith 1.5 35B A3Bornith-ai36B MoE21.4 GB36.2 GBGQA256KAug 18, 2026
Kimi K2.7 Codemoonshotai1.03T MoE579 GB1017 GBMLA256KJun 11, 2026
gemma 3 27b itgoogle27B17.3 GB28.7 GBGQA + sliding window128KMar 1, 2025
Phi 3 mini 4k instructmicrosoft3.8B5.7 GB7.2 GBMHA4KApr 22, 2024
Llama 3.2 3Bmeta-llama3.2B3.6 GB5.1 GBGQA128KSep 18, 2024
DeepSeek V4 Prodeepseek-ai1.60T MoE891 GB1559 GBGQA1MApr 22, 2026
GLM 5.2zai-org753B MoE435 GB748 GBMLA1MJun 16, 2026
Phi 4 mini instructmicrosoft3.8B4.1 GB5.6 GBGQA128KFeb 19, 2025
DeepSeek V3.1deepseek-ai685B MoE379 GB665 GBMLA160KAug 21, 2025
DeepSeek R1 Distill Qwen 7Bdeepseek-ai7.6B5.5 GB8.7 GBGQA128KJan 20, 2025
gemma 3 4b ptgoogle4.3B3.6 GB5.4 GBGQA + sliding window128KFeb 20, 2025
Phi 3.5 vision instructmicrosoft4.1B5.8 GB7.6 GBMHA128KAug 16, 2024
Qwen3.5 9B BaseQwen9.7B6.4 GB10.0 GBGQA256KFeb 26, 2026
SmolVLM 500M InstructHuggingFaceTB507M1.1 GB1.2 GBGQA8KJan 20, 2025
DeepSeek OCR 2deepseek-ai3.4B MoE2.9 GB4.0 GBMHA8KJan 27, 2026
SmolLM2 1.7B InstructHuggingFaceTB1.7B3.0 GB3.7 GBMHA8KOct 31, 2024
Kimi K2 Instructmoonshotai1.03T MoE579 GB1017 GBMLA128KJul 11, 2025
Phi 3 mini 128k instructmicrosoft3.8B5.7 GB7.2 GBMHA128KApr 22, 2024
medgemma 1.5 4b itgoogle4.3B3.5 GB5.0 GBGQA + sliding window128KJan 7, 2026
UI TARS 1.5 7BByteDance-Seed8.3B5.5 GB8.7 GBGQA125KApr 16, 2025
SmolLM2 360M InstructHuggingFaceTB362M1.1 GB1.2 GBGQA8KOct 31, 2024
MiniMax M2.7MiniMaxAI229B MoE133 GB229 GBGQA200KApr 9, 2026
SmolLM3 3BHuggingFaceTB3.1B3.0 GB4.3 GBGQA64KJul 8, 2025
phi 4microsoft15B10.4 GB16.7 GBGQA16KDec 11, 2024
Kimi K2.6moonshotai1.03T MoE586 GB1017 GBMLA256KApr 14, 2026
Apertus 8B Instruct 2509swiss-ai8.1B6.4 GB9.6 GBGQA64KAug 13, 2025
DeepSeek V4 Flash DSparkdeepseek-ai304B MoE175 GB302 GBGQA1MJun 27, 2026
Step 3.5 Flashstepfun-ai199B116 GB199 GBMHA + sliding window256KFeb 1, 2026
phi 2microsoft2.8B4.7 GB5.8 GBMHA2KDec 13, 2023
Llama Guard 4 12Bmeta-llama12B9.1 GB14.2 GBGQA10MApr 23, 2025
LFM2.5 350MLiquidAI354M0.8 GB1.0 GBGQA125KMar 31, 2026
Muse Glimmer 30Bmeta-models30B17.9 GB28.5 GBGQA + sliding window128KAug 9, 2026
DeepSeek V3 0324deepseek-ai685B MoE378 GB665 GBMLA160KMar 24, 2025
DeepSeek R1 0528deepseek-ai685B MoE378 GB665 GBMLA160KMay 28, 2025
SmolLM2 360MHuggingFaceTB362M1.1 GB1.2 GBGQA8KOct 31, 2024
Olmo 3 7B Instructallenai7.3B7.3 GB10.3 GBMHA + sliding window64KNov 19, 2025
Meta Llama 3.1 8B InstructNousResearch8.0B6.2 GB9.6 GBGQA128KJul 24, 2024
GOT OCR2 0stepfun-ai716M1.8 GB2.1 GBMHA32KSep 12, 2024
MiniMax M3MiniMaxAI427B MoE248 GB424 GBGQA1MJun 2, 2026
DeepSeek R1 Distill Qwen 14Bdeepseek-ai15B10.6 GB16.8 GBGQA128KJan 20, 2025
falcon 7btiiuae7.2B9.6 GB12.1 GBMHA—Apr 24, 2023
GLM 4.5zai-org358B MoE205 GB358 GBGQA128KJul 20, 2025
OLMo 2 0425 1Ballenai1.5B2.5 GB3.1 GBMHA4KApr 17, 2025
DeepSeek V2 Litedeepseek-ai16B MoE9.8 GB16.4 GBMLA160KMay 15, 2024
GOT OCR 2.0 hfstepfun-ai561M1.8 GB2.0 GBMHA—Nov 22, 2024
granite 3.1 3b a800m instructibm-granite3.3B MoE2.9 GB4.3 GBGQA128KDec 6, 2024
Llama 4 Scout 17B 16E Instructmeta-llama109B MoE63.2 GB109 GBGQA10MApr 2, 2025
MiniMax M2MiniMaxAI229B MoE132 GB229 GBGQA192KOct 22, 2025
Mistral 7B Instruct v0.1mistralai7.2B5.5 GB8.6 GBGQA32KSep 27, 2023
LFM3 1.7BLiquidAI1.7B1.8 GB2.5 GBGQA + sliding window256KOct 23, 2026
LFM2.5 1.2B InstructLiquidAI1.2B1.3 GB1.8 GBGQA125KJan 6, 2026
Olmo 3 1025 7Ballenai7.3B7.3 GB10.3 GBMHA + sliding window64KSep 12, 2025
granite 3.3 8b instructibm-granite8.2B6.3 GB9.8 GBGQA128KApr 9, 2025
PaddleOCR VL 1.6PaddlePaddle959M1.3 GB1.7 GBGQA128KMay 27, 2026
Hy3 previewtencent299B MoE173 GB299 GBGQA256KApr 13, 2026
deepseek coder 7b instruct v1.5deepseek-ai6.9B8.3 GB11.2 GBMHA4KJan 25, 2024
SmolLM3 3B BaseHuggingFaceTB3.1B3.0 GB4.3 GBGQA64KJun 19, 2025
Llama 2 7b hfNousResearch6.7B8.3 GB11.1 GBMHA4KJul 18, 2023
DeepSeek Coder V2 Lite Instructdeepseek-ai16B MoE10.5 GB16.4 GBMLA160KJun 14, 2024
ERNIE 4.5 21B A3B Thinkingbaidu22B13.5 GB22.6 GBGQA128KSep 8, 2025
Meta Llama 3 8B InstructNousResearch8.0B6.2 GB9.6 GBGQA8KApr 18, 2024
granite vision 4.1 4bibm-granite4.0B3.2 GB4.6 GBGQA128KApr 16, 2026
granite 4.1 8bibm-granite8.8B6.8 GB10.5 GBGQA128KApr 6, 2026
olmOCR 2 7B 1025allenai8.3B5.5 GB8.7 GBGQA125KOct 6, 2025
MiMo V2.5XiaomiMiMo311B MoE180 GB308 GBGQA1MApr 27, 2026
DeepSeek R1 Distill Llama 8Bdeepseek-ai8.0B6.2 GB9.6 GBGQA128KJan 20, 2025
Hunyuan A13B Instructtencent80B MoE47.1 GB81.2 GBGQA32KJun 25, 2025
deepseek vl2 tinydeepseek-ai3.4B MoE3.0 GB4.5 GBMHA4KDec 13, 2024
MiniMax M3.1 FlashMiniMaxAI220B MoE127 GB220 GBGQA1MOct 17, 2026
Olmo 3 7B Thinkallenai7.3B7.3 GB10.3 GBMHA + sliding window64KNov 18, 2025
DeepSeek V2 Lite Chatdeepseek-ai16B MoE10.5 GB16.4 GBMLA160KMay 15, 2024
granite 4.2 8bibm-granite8.8B6.8 GB10.5 GBGQA128KAug 7, 2026
flintRecursiveSI40B MoE23.6 GB40.3 GBGQA + sliding window256KOct 21, 2026
Kimi K2 Thinkingmoonshotai1.03T MoE580 GB1017 GBMLA256KNov 4, 2025
DeepSeek V3.2 Expdeepseek-ai685B MoE391 GB680 GBMLA160KSep 29, 2025
GLM 4.6V Flashzai-org10B6.7 GB10.3 GBGQA128KDec 7, 2025
granite 3.2 8b instructibm-granite8.2B6.3 GB9.8 GBGQA128KFeb 17, 2025
Llama 3.2 11B Vision Instructmeta-llama11B7.4 GB11.6 GBGQA128KSep 18, 2024
DeepSeek V4 Pro 0813deepseek-ai1.65T MoE941 GB1635 GBGQA1MAug 13, 2026
GLM 4.7zai-org358B MoE205 GB358 GBGQA198KDec 22, 2025
deepseek coder 6.7b basedeepseek-ai6.7B8.3 GB11.1 GBMHA16KOct 23, 2023
Kimi K3.1 Promoonshotai2.78T MoE1584 GB2753 GBMLA1MNov 2, 2026
internlm2 chat 7binternlm7.7B6.0 GB9.2 GBGQA32KJan 10, 2024
Kimi Linear 48B A3B Instructmoonshotai49B MoE28.9 GB49.6 GBMLA—Oct 30, 2025
internlm3 8b instructinternlm8.8B6.0 GB9.7 GBGQA32KJan 13, 2025
Molmo2 8Ballenai8.7B6.5 GB9.9 GBGQA36KDec 14, 2025
OLMoE 1B 7B 0924allenai6.9B MoE5.4 GB8.4 GBMHA4KJul 20, 2024
step3stepfun-ai321B214 GB349 GBMHA64KJul 28, 2025
Seed OSS 36B InstructByteDance-Seed36B23.0 GB38.5 GBGQA512KAug 20, 2025
SmolLM2 1.7BHuggingFaceTB1.7B3.0 GB3.7 GBMHA8KOct 30, 2024
LFM2.5 VL 450MLiquidAI449M0.8 GB1.0 GBGQA125KApr 8, 2026
MiMo V2.6 Flash RLXiaomiMiMo311B MoE179 GB309 GBGQA1MSep 21, 2026
granite 4.2 3bibm-granite3.7B3.3 GB4.8 GBGQA128KAug 7, 2026
granite 3.3 2b instructibm-granite2.5B2.6 GB3.6 GBGQA128KApr 9, 2025
Kimi VL A3B Instructmoonshotai16B MoE10.8 GB16.8 GBMLA128KApr 9, 2025
GLM 4.1V 9B Thinkingzai-org10B6.7 GB10.3 GBGQA64KJun 28, 2025
olmOCR 7B 0225 previewallenai8.3B5.5 GB8.7 GBGQA32KJan 15, 2025
granite 3.1 8b instructibm-granite8.2B6.3 GB9.8 GBGQA128KDec 6, 2024
granite 4.2 30bibm-granite29B19.1 GB31.6 GBGQA128KAug 7, 2026
GLM Z1 32B 0414zai-org33B19.5 GB33.4 GBGQA32KApr 8, 2025
LFM2.5 2.6BLiquidAI2.7B2.3 GB3.4 GBGQA128KJul 28, 2026
falcon 7b instructtiiuae7.2B9.4 GB11.8 GBMHA—Apr 25, 2023
DeepSeek V4 Flash Vision Expdeepseek-ai305B MoE175 GB303 GBGQA1MAug 31, 2026
granite 4.0 1bibm-granite1.6B2.2 GB2.8 GBGQA128KOct 7, 2025
deepseek coder 1.3b instructdeepseek-ai1.3B2.8 GB3.3 GBMHA16KOct 29, 2023
Moonlight 16B A3B Instructmoonshotai16B MoE10.8 GB16.8 GBMLA8KFeb 22, 2025
MiniMax M1 40kMiniMaxAI456B MoE263 GB455 GBGQA9.765625MJun 5, 2025
LFM2.5 24B A2BLiquidAI24B MoE14.3 GB24.5 GBGQA128KOct 12, 2026
MiMo 7B RLXiaomiMiMo7.8B6.2 GB9.4 GBGQA32KApr 29, 2025
deepseek coder 6.7b instructdeepseek-ai6.7B8.3 GB11.1 GBMHA16KOct 29, 2023
granite 4.0 h tinyibm-granite6.9B MoE4.6 GB7.5 GBGQA128KSep 16, 2025
Yi 34B Chat01-ai34B21.6 GB36.4 GBGQA4KNov 22, 2023
Qwen3.8 2.4T A95BQwen2.45T MoE1394 GB2423 GBGQA256KAug 8, 2026
gpt oss safeguard 20bopenai22B MoE11.8 GB12.3 GBGQA + sliding window128KSep 18, 2025
LLaDA2.1 miniinclusionAI16B MoE10.3 GB17.1 GBGQA32KFeb 9, 2026
Molmo2 4Ballenai4.9B4.3 GB6.2 GBGQA36KDec 14, 2025
GLM 4.6zai-org357B MoE204 GB357 GBGQA198KSep 29, 2025
granite 4.0 h smallibm-granite32B MoE18.9 GB32.6 GBGQA128KSep 16, 2025
WeVisOne VL 4Btencent4.8B4.3 GB6.2 GBGQA256KOct 22, 2026
DeepSeek V2 Chatdeepseek-ai236B MoE135 GB234 GBMLA160KApr 28, 2024
LFM3 VL 2BLiquidAI1.9B1.8 GB2.5 GBGQA + sliding window125KOct 19, 2026
deepseek moe 16b basedeepseek-ai16B MoE12.5 GB18.6 GBMHA4KJan 8, 2024
Nemotron 3.5 Parsenvidia631M1.1 GB1.5 GBGQA16KSep 3, 2026
MiMo V2.6 Distill Qwen 9BXiaomiMiMo9.4B6.6 GB10.0 GBGQA256KSep 21, 2026
SmolLM 1.7BHuggingFaceTB1.7B3.0 GB3.7 GBMHA2KJul 14, 2024
MiMo V2.6 Pro RLXiaomiMiMo1.02T MoE587 GB1017 GBGQA1MSep 21, 2026
DeepSeek R1 Distill Llama 70Bdeepseek-ai71B43.3 GB73.0 GBGQA128KJan 20, 2025
glm 4 9bzai-org9.4B11.5 GB15.0 GBMHA8KJun 4, 2024
MiMo V2.6 Flash MOPDXiaomiMiMo311B MoE179 GB309 GBGQA1MSep 27, 2026
Intern S1internlm241B MoE139 GB235 GBGQA64KJul 24, 2025
d1 3BLiquidAI3.1B2.3 GB3.4 GBGQA32KOct 5, 2026
granite 4.0 microibm-granite3.4B3.2 GB4.6 GBGQA128KSep 16, 2025
FastVLM 0.5Bapple759M1.2 GB1.5 GBGQA32KAug 25, 2025
Llama 3.1 70Bmeta-llama71B42.7 GB73.0 GBGQA128KJul 14, 2024
gpt oss safeguard 120bopenai117B MoE59.5 GB60.1 GBGQA + sliding window128KSep 18, 2025
Qianfan OCRbaidu4.7B4.4 GB6.2 GBGQA32KMar 18, 2026
Yi 6B 200K01-ai6.1B4.4 GB7.0 GBGQA195KNov 6, 2023
LFM2.5 VL 3BLiquidAI3.1B2.3 GB3.4 GBGQA32KAug 11, 2026
internlm2 5 7b chatinternlm7.7B6.0 GB9.2 GBGQA32KJun 27, 2024
SmolVLM InstructHuggingFaceTB2.2B3.0 GB3.8 GBMHA16KNov 18, 2024
Inkling 2 Smallthinkingmachines594B MoE346 GB590 GBGQA—Nov 4, 2026
ERNIE 4.5 21B A3B PTbaidu22B13.5 GB22.6 GBGQA128KJun 28, 2025
LFM2.5 8B A1BLiquidAI8.5B MoE5.5 GB9.1 GBGQA125KMay 28, 2026
Yi 6B01-ai6.1B4.4 GB7.0 GBGQA4KNov 1, 2023
WeVisOne VL 2Btencent2.4B2.8 GB3.6 GBGQA256KOct 22, 2026
MiMo V2 FlashXiaomiMiMo310B MoE176 GB308 GBGQA256KDec 16, 2025
LFM2 350MLiquidAI354M1.0 GB1.1 GBGQA125KJul 10, 2025
Yi 1.5 34B Chat01-ai34B21.6 GB36.4 GBGQA4KMay 10, 2024
GLM 4.5Vzai-org108B MoE61.4 GB108 GBGQA64KAug 10, 2025
LongCat Flash Chatmeituan-longcat562B MoE321 GB557 GBMLA128KAug 29, 2025
Muse Spark Open 816B A42Bmeta-models822B MoE470 GB813 GBGQA + sliding window1MSep 15, 2026
LightOnOCR 3 4Blightonai4.5B3.6 GB5.3 GBGQA256KOct 6, 2026
LightOnOCR 3 0.8Blightonai853M1.5 GB1.7 GBGQA256KOct 7, 2026
d1 2BLiquidAI1.9B1.8 GB2.5 GBGQA + sliding window125KOct 24, 2026
Kimi K2 Basemoonshotai1.03T MoE586 GB1017 GBMLA128KJul 3, 2025
Ling 3.0 tinyinclusionAI7.9B MoE5.5 GB8.7 GBMLA128KAug 10, 2026
MiniMax M2.1MiniMaxAI229B MoE132 GB229 GBGQA192KDec 20, 2025
Llama 4 Maverick 17B 128E Instructmeta-llama402B MoE228 GB399 GBGQA1MApr 1, 2025
Kimi VL A3B Thinking 2506moonshotai16B MoE10.8 GB16.8 GBMLA128KJun 21, 2025
LFM3 1.7B BaseLiquidAI1.7B1.7 GB2.5 GBGQA + sliding window256KOct 26, 2026
Yi 34B01-ai34B21.6 GB36.4 GBGQA4KNov 1, 2023
Trinity Miniarcee-ai26B MoE15.8 GB26.9 GBGQA + sliding window128KDec 1, 2025
Hy4 previewtencent780B MoE445 GB773 GBMLA1MAug 27, 2026
falcon 40btiiuae42B25.2 GB42.9 GBGQA—May 24, 2023
Nous Hermes 2 Mixtral 8x7B DPONousResearch47B MoE28.0 GB47.7 GBGQA32KJan 11, 2024
Yi 34B 200K01-ai34B21.6 GB36.4 GBGQA195KNov 6, 2023
Step 3.7 Flashstepfun-ai201B119 GB203 GBMHA + sliding window256KMay 23, 2026
Nous Hermes 2 Yi 34BNousResearch34B22.1 GB36.4 GBGQA4KDec 23, 2023
Kimi K2 Instruct 0905moonshotai1.03T MoE580 GB1017 GBMLA256KSep 3, 2025
Arete 27BNousResearch27B17.4 GB28.1 GBGQA256KNov 6, 2026
Qwen AgentWorld 35B A3BQwen35B MoE21.6 GB35.4 GBGQA256KJun 22, 2026
MiMo V2.5 ProXiaomiMiMo1.02T MoE590 GB1017 GBGQA1MApr 27, 2026
Ling 3.0 flashinclusionAI127B MoE73.5 GB127 GBMLA256KAug 2, 2026
Kimi VL A3B Thinkingmoonshotai16B MoE10.3 GB17.2 GBMLA128KApr 9, 2025
North Mini Code 1.0CohereLabs30B MoE19.3 GB31.6 GBGQA + sliding window488KJun 5, 2026
GLM 4 32B 0414zai-org33B19.5 GB33.4 GBGQA32KApr 7, 2025
Hy Memory 30B A3Btencent30B MoE18.5 GB31.1 GBGQA256KOct 15, 2026
Llama 3.2 11B Visionmeta-llama11B8.0 GB12.4 GBGQA128KSep 18, 2024
Ling 3.1 flashinclusionAI572B MoE335 GB568 GBMLA256KOct 13, 2026
GLM 4.6Vzai-org108B MoE67.8 GB108 GBGQA128KDec 7, 2025
Qwen3.8 27B DSpine b16tencent2.0B2.2 GB3.1 GBGQA + sliding window256KNov 7, 2026
Intern Decision 4Binternlm4.5B3.6 GB5.3 GBGQA256KSep 26, 2026
Mage VL Coworkmicrosoft4.7B4.3 GB6.2 GBGQA256KOct 30, 2026
Step3 VL 10Bstepfun-ai10B6.5 GB9.9 GBGQA64KJan 13, 2026
Ling 3.1 flash 1019inclusionAI572B MoE335 GB568 GBMLA256KOct 27, 2026
SenseNova Loop 9B Previewsensenova9.7B6.4 GB10.0 GBGQA256KOct 28, 2026
Hermes 3 Llama 3.2 3BNousResearch3.2B3.4 GB4.7 GBGQA128KDec 3, 2024
Ling 1TinclusionAI1000B MoE566 GB993 GBGQA32KOct 2, 2025
ERNIE 5.5 Flash Basebaidu286B MoE164 GB284 GBMLA1MOct 21, 2026
Kimi Dev 72Bmoonshotai73B44.6 GB75.1 GBGQA128KJun 16, 2025
MiniMax Text 01MiniMaxAI456B MoE263 GB455 GBGQA9.765625MJan 12, 2025
GLiDE 27Bfastino26B16.2 GB27.2 GBGQA256KOct 19, 2026
Falcon H1R 7Btiiuae7.6B5.3 GB8.5 GBGQA256KOct 29, 2025
Hermes 4.3 36BNousResearch36B23.0 GB38.5 GBGQA512KNov 17, 2025
LongCat 2.0meituan-longcat1.78T MoE1012 GB1758 GBMLA256KJul 5, 2026
Ling 3.0 flash SanteinclusionAI127B MoE75.9 GB127 GBMLA256KOct 19, 2026
MiniMax H3 Image PEMiniMaxAI27B16.8 GB28.0 GBGQA256KNov 8, 2026
UI TARS 7B DPOByteDance-Seed8.3B5.5 GB8.7 GBGQA32KJan 22, 2025
Groot 19B A3B 261023GrootTeam19B MoE12.7 GB19.9 GBGQA256KOct 23, 2026
granite 4.1 3b llmon previewibm-granite3.7B3.3 GB4.8 GBGQA128KOct 1, 2026
S5 nanomaritaca-ai4.7B3.7 GB5.4 GBGQA256KNov 2, 2026
Meta Llama Guard 2 8Bmeta-llama8.0B6.2 GB9.6 GBGQA8KApr 17, 2024
Axion 1.0 35BAxion1-035B MoE20.7 GB35.8 GBGQA256KOct 16, 2026
LongCat 2.5meituan-longcat1.78T MoE1014 GB1762 GBMLA512KNov 4, 2026
Llama 4 Scout 17B 16Emeta-llama109B MoE64.1 GB110 GBGQA256KApr 2, 2025
Raon 1 nanoKRAFTON3.5B3.0 GB4.3 GBMLA256KOct 29, 2026
SingVL S1 4BinclusionAI4.5B3.6 GB5.3 GBGQA256KNov 11, 2026
Nex N2.6nex-agi321B MoE188 GB319 GBMLA1MNov 10, 2026
Maxwell V1 Flashmaxwell-model37B MoE22.1 GB37.6 GBGQA512KOct 13, 2026