智能指数
综合全部模型按智能指数排序，共 582 个，可上下滚动
1
Claude Opus 5 (Adaptive Reasoning, Max Effort)63.12Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)62.53Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)62.14Claude Opus 5 (Adaptive Reasoning, High Effort)61.55
GPT-5.6 Sol (max)60.96
Kimi K3 (max)59.77GPT-5.6 Sol (xhigh)59.08Claude Opus 5 (Adaptive Reasoning, Medium Effort)58.69
Qwen3.8 Max58.110GPT-5.6 Sol (high)57.311Claude Opus 4.8 (Adaptive Reasoning, Max Effort)57.312
Muse Spark 1.2 (xhigh)56.813GPT-5.6 Terra (max)56.614GPT-5.5 (xhigh)56.315
Grok 4.5 (high)55.816GPT-5.6 Sol (medium)55.617Claude Sonnet 5 (Adaptive Reasoning, Max Effort)55.318Claude Opus 4.7 (Adaptive Reasoning, Max Effort)55.019GPT-5.5 (high)54.720Muse Spark 1.1 (xhigh)53.221GPT-5.4 (xhigh)53.122GPT-5.6 Terra (xhigh)52.823
GLM-5.2 (max)52.624Claude Opus 5 (Adaptive Reasoning, Low Effort)52.525GPT-5.6 Luna (max)52.326
Gemini 3.5 Flash (high)52.027
DeepSeek V4 Flash 0731 (Reasoning, Max Effort)51.828Gemini 3.6 Flash (high)51.629GPT-5.5 (medium)51.430GPT-5.6 Sol (low)50.731GPT-5.6 Terra (high)50.132GPT-5.6 Luna (xhigh)50.133Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)48.434Kimi K3 (low)48.335Gemini 3.1 Pro Preview47.736GPT-5.6 Luna (high)47.037GPT-5.6 Terra (medium)46.838Gemini 3.5 Flash (medium)46.739Qwen3.7 Max46.740GPT-5.3 Codex (xhigh)45.541
MiniMax-M345.442DeepSeek V4 Pro (Reasoning, Max Effort)45.343Kimi K2.645.144
Motif 3 (Beta)44.945Claude Opus 4.6 (Adaptive Reasoning, Max Effort)44.946GPT-5.5 (low)44.547Claude Opus 4.7 (Non-reasoning, High Effort)43.948DeepSeek V4 Pro (Reasoning, High Effort)43.749Muse Spark43.550GPT-5.2 (xhigh)43.351Kimi K2.7 Code43.052
MiMo-V2.5-Pro42.953Claude Sonnet 5 (Non-reasoning, High Effort)42.654
Inkling (xhigh)42.355
Hy342.256DeepSeek V4 Flash (Reasoning, Max Effort)42.157GPT-5.6 Sol (Non-reasoning)41.958Claude Opus 4.5 (Reasoning)41.959
Nex-N2-Pro41.760MiMo-V2-Pro41.461GPT-5.6 Terra (low)41.362Inkling Small41.263GPT-5.2 Codex (xhigh)41.264Qwen3.6 Max Preview41.165GLM-5.1 (Reasoning)41.066GPT-5.4 mini (xhigh)40.967Grok Build 0.1 061640.768Gemini 3 Pro Preview (high)40.669GLM-5 (Reasoning)40.670Qwen3.6 Plus40.571GPT-5.4 (low)40.272
JT-4.1 Flash 236B A21B39.873
Agnes 2.5 Pro Alpha39.774GPT-5.4 nano (xhigh)39.775Qwen3.7 Plus39.476GLM-5-Turbo39.177DeepSeek V4 Flash (Reasoning, High Effort)39.078GPT-5.6 Luna (medium)38.979GPT-5.2 (medium)38.980MiniMax-M2.738.981Claude Opus 4.6 (Non-reasoning, High Effort)38.882Gemini 3 Flash Preview (Reasoning)38.783
Nemotron 3 Ultra 550B A55B (Reasoning)38.384MiMo-V2.538.085Grok 4.20 0309 v2 (Reasoning)38.086Grok 4.3 (high)37.987
Ling-3.0-flash37.888Qwen3.6 27B (Reasoning)37.789GPT-5.1 (high)37.590Gemini 3.5 Flash-Lite37.491Claude 4.5 Sonnet (Reasoning)37.492Grok 4.20 0309 (Reasoning)37.493MiMo-V2-Omni-032737.394GPT-5 Codex (high)37.095Grok 4.3 (medium)36.996Claude Sonnet 4.6 (Non-reasoning, High Effort)36.897Grok 4.3 (low)36.398GLM-5.1 (Non-reasoning)36.399Kimi K2.5 (Reasoning)36.0100MiMo-V2-Omni35.9101Gemini 3.5 Flash (minimal)35.8102GPT-5.5 (Non-reasoning)35.8103GPT-5.1 Codex (high)35.6104Claude Opus 4.5 (Non-reasoning)35.6105Kimi K2.6 (Non-reasoning)35.4106GPT-5 (high)35.3107GLM 5V Turbo (Reasoning)35.3108Claude Sonnet 4.6 (Non-reasoning, Low Effort)35.1109GLM-5.2 (Non-reasoning)34.8110GPT-5.6 Terra (Non-reasoning)34.6111GPT-5 (medium)34.6112Qwen3.5 27B (Reasoning)34.6113Claude 4.1 Opus (Reasoning)34.5114MiniMax-M2.534.5115GLM-4.7 (Reasoning)34.5116Hy3-preview (Reasoning)34.4117Qwen3.5 397B A17B (Reasoning)34.3118GPT-5.5 Instant (May 2026)34.3119Grok 434.1120MiMo-V2-Flash (Feb 2026)34.0121GPT-5.6 Luna (low)33.9122
KAT Coder Pro V233.9123
LongCat 2.033.9124Gemini 3 Pro Preview (low)33.9125Kimi K2 Thinking33.5126o3-pro33.3127GLM-5 (Non-reasoning)33.2128Qwen3.5 122B A10B (Reasoning)32.8129Qwen3.5 397B A17B (Non-reasoning)32.7130DeepSeek V3.2 (Reasoning)32.6131Qwen3 Max Thinking32.5132Qwen3.6 35B A3B (Reasoning)32.1133MiniMax-M2.132.1134DeepSeek V4 Pro (Non-reasoning)31.9135GPT-5 (low)31.9136MiMo-V2-Flash (Reasoning)31.9137Claude 4 Opus (Reasoning)31.7138GPT-5 mini (medium)31.6139
G9v3-39A5B31.4140Qwen3.6 27B (Non-reasoning)31.3141Qwen3.5 Omni Plus31.3142GPT-5.1 Codex mini (high)31.3143Grok 4.1 Fast (Reasoning)31.3144o331.1145Ring-2.6-1T31.1146
Step 3.7 Flash30.9147GPT-5.4 nano (medium)30.8148DeepSeek V3.1 Terminus (Reasoning)30.6149GPT-5.4 mini (medium)30.5150
Mistral Medium 3.530.4151Kimi K2.5 (Non-reasoning)30.1152Qwen3.5 27B (Non-reasoning)30.0153Claude 4.5 Haiku (Reasoning)29.9154Claude 4.5 Sonnet (Non-reasoning)29.9155Qwen3.5 35B A3B (Reasoning)29.9156Claude 4 Sonnet (Reasoning)29.8157Gemma 4 31B (Reasoning)29.7158DeepSeek V4 Flash (Non-reasoning)29.3159GLM-4.6 (Reasoning)29.3160GPT-5.5 Instant (June 2026)29.2161JT-35B-Flash29.0162KAT-Coder-Pro V128.9163MiniMax-M228.9164Claude 4.1 Opus (Non-reasoning)28.8165MiMo-V2.5-Pro (Non-reasoning)28.4166GPT-5.4 (Non-reasoning)28.3167Qwen3.5 122B A10B (Non-reasoning)28.2168Gemini 3 Flash Preview (Non-reasoning)27.9169Grok 4 Fast (Reasoning)27.9170Claude 3.7 Sonnet (Reasoning)27.6171GLM-4.7 (Non-reasoning)27.1172GPT-5.6 Luna (Non-reasoning)26.8173Hy3-preview (Non-reasoning)26.6174Ling-2.6-1T26.6175
Doubao Seed Code26.5176GPT-5.2 (Non-reasoning)26.5177Step 3.5 Flash 260326.5178Gemma 4 26B A4B (Reasoning)26.1179o4-mini (high)26.1180Claude 4 Opus (Non-reasoning)26.0181Claude 4 Sonnet (Non-reasoning)26.0182Step 3.5 Flash26.0183Gemini 2.5 Pro25.9184DeepSeek V3.2 Exp (Reasoning)25.9185GPT-5 mini (high)25.8186NVIDIA Nemotron 3 Super 120B A12B (Reasoning)25.7187Gemini 3.1 Flash-Lite25.6188Qwen3 Max Thinking (Preview)25.5189DeepSeek V3.2 (Non-reasoning)25.1190Grok 4.3 (Non-reasoning)25.0191MiMo-V2-Flash (Non-reasoning)25.0192Qwen3.6 35B A3B (Non-reasoning)24.5193Qwen3 Max24.5194Qwen3.5 35B A3B (Non-reasoning)24.3195Ling 3.0 Tiny24.3196Gemini 2.5 Flash Preview (Sep '25) (Reasoning)24.2197gpt-oss-120b (high)24.1198Claude 4.5 Haiku (Non-reasoning)24.1199Kimi K2 090524.0200o123.9201Claude 3.7 Sonnet (Non-reasoning)23.9202Gemini 2.5 Pro Preview (Mar' 25)23.4203GLM-4.6 (Non-reasoning)23.4204GLM-4.7-Flash (Reasoning)23.3205Grok 3 mini Reasoning (high)22.9206Grok 4.20 0309 (Non-reasoning)22.9207
Command A+22.8208Gemini 2.5 Pro Preview (May' 25)22.7209DeepSeek V3.2 Speciale22.6210
K-EXAONE (Reasoning)22.4211Gemma 4 31B (Non-reasoning)22.3212Gemma 4 12B (Reasoning)22.3213
ERNIE 5.0 Thinking Preview22.3214Grok 4.20 0309 v2 (Non-reasoning)22.2215
Nova 2.0 Pro Preview (medium)22.0216Grok Code Fast 122.0217
Mercury 221.9218Qwen3.5 9B (Reasoning)21.8219DeepSeek V3.1 Terminus (Non-reasoning)21.7220DeepSeek V3.2 Exp (Non-reasoning)21.7221
Apriel-v1.5-15B-Thinker21.6222DeepSeek V3.1 (Non-reasoning)21.4223Nova 2.0 Omni (medium)21.3224Qwen3 Coder Next21.3225DeepSeek V3.1 (Reasoning)21.0226Qwen3 VL 235B A22B (Reasoning)20.9227Apriel-v1.6-15B-Thinker20.8228GPT-5.1 (Non-reasoning)20.7229Qwen3.5 9B (Non-reasoning)20.6230EXAONE 4.5 33B20.5231Gemma 4 26B A4B (Non-reasoning)20.4232Qwen3.5 4B (Reasoning)20.4233DeepSeek R1 0528 (May '25)20.4234Gemini 2.5 Flash (Reasoning)20.3235North Mini Code20.2236GPT-5 nano (high)20.1237Nova 2.0 Pro Preview (low)20.0238Qwen3 235B A22B 2507 (Reasoning)19.9239Mistral Small 4 (Reasoning)19.7240Kimi K219.7241GLM-4.5 (Reasoning)19.7242GPT-4.119.6243Qwen3 Max (Preview)19.4244Devstral 219.2245Nova 2.0 Lite (medium)19.2246Qwen3.5 Omni Flash19.2247GPT-5 nano (medium)19.2248o3-mini19.2249o1-pro19.1250Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning)19.1251JT-MINI18.8252
Trinity Large Thinking18.6253DeepSeek R1 (Jan '25)18.6254Grok 318.6255Seed-OSS-36B-Instruct18.5256Nova 2.0 Lite (high)18.4257Qwen3 235B A22B 2507 Instruct18.4258
HyperNova 60B 260518.3259Qwen3 Coder 480B A35B Instruct18.2260Qwen3 VL 32B (Reasoning)18.1261Magistral Medium 1.218.0262Nova 2.0 Lite (low)18.0263Nemotron Cascade 2 30B A3B18.0264
Sonar Reasoning Pro18.0265MiniMax M1 80k17.9266GPT-5.4 nano (Non-Reasoning)17.8267Devstral Small 217.7268Gemini 2.5 Flash Preview (Reasoning)17.7269LongCat Flash Lite17.4270GPT-5 (minimal)17.3271
K2 Think V217.2272
HyperCLOVA X SEED Think (32B)17.2273o1-preview17.2274Grok 4.1 Fast (Non-reasoning)17.0275K-EXAONE (Non-reasoning)16.9276Qwen3 Next 80B A3B (Reasoning)16.9277GLM-4.6V (Reasoning)16.9278GPT-5.4 mini (Non-Reasoning)16.8279Nova 2.0 Omni (low)16.7280GLM-4.5-Air16.7281
Mi:dm K 2.5 Pro16.6282Grok 4 Fast (Non-reasoning)16.6283G9v3-3B16.5284Ring-1T16.3285Qwen3.5 4B (Non-reasoning)16.1286Mistral Large 315.9287
INTELLECT-315.7288o3-mini (high)15.7289GLM-4.7-Flash (Non-reasoning)15.6290GPT-5 (ChatGPT)15.4291gpt-oss-20b (high)15.2292
Solar Open 100B (Reasoning)15.2293Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning)15.2294DeepSeek V3 032415.2295Grok 3 Reasoning Beta15.2296Nemotron 3 Nano Omni 30B A3B Reasoning15.0297gpt-oss-120b (low)14.9298Mistral Small 3.114.9299GPT-4.1 mini14.8300Mistral Medium 3.114.7301Qwen3 30B A3B 2507 (Reasoning)14.6302Llama 4 Maverick14.5303Solar Pro 314.5304NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)14.5305MiniMax M1 40k14.5306gpt-oss-20b (low)14.4307Nova 2.0 Pro Preview (Non-reasoning)14.4308Qwen3 VL 235B A22B Instruct14.4309GPT-5 mini (minimal)14.3310K2-V2 (high)14.2311Ling 2.6 Flash14.2312Gemini 2.5 Flash (Non-reasoning)14.2313DeepSeek V3 (Dec '24)14.2314o1-mini14.0315DiffusionGemma 26B A4B13.8316Qwen3 Next 80B A3B Instruct13.8317GPT-4.5 (Preview)13.6318
Tri-21B-think Preview13.6319Qwen3 Coder 30B A3B Instruct13.6320Qwen3 235B A22B (Reasoning)13.5321QwQ 32B13.4322Qwen3 VL 30B A3B (Reasoning)13.4323Gemini 2.0 Flash Thinking Experimental (Jan '25)13.3324Gemma 4 12B (Non-reasoning)13.2325Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)13.1326Motif-2-12.7B-Reasoning12.8327Nova Premier12.7328Ling-1T12.7329
Celeris-112.5330Magistral Medium 112.5331Mistral Medium 312.5332Solar Pro 2 (Preview) (Reasoning)12.5333Gemma 4 E4B (Reasoning)12.4334Llama Nemotron Super 49B v1.5 (Reasoning)12.4335K2-V2 (medium)12.4336Claude 3.5 Haiku12.4337Devstral Medium12.4338Mistral Small 4 (Non-reasoning)12.3339Tri-21B-Think12.3340GPT-4o (March 2025, chatgpt-4o-latest)12.3341Gemini 2.0 Flash (Feb '25)12.2342Llama 3.3 Nemotron Super 49B v1 (Reasoning)12.2343
Sarvam 105B (high)11.9344
MiniCPM5-1B (Reasoning)11.9345Qwen3 4B 2507 (Reasoning)11.9346Nova 2.0 Lite (Non-reasoning)11.8347Gemini 2.0 Pro Experimental (Feb '25)11.8348Claude 3 Opus11.8349Devstral Small (May '25)11.8350MiniCPM5-1B (Non-reasoning)11.7351Gemini 2.5 Flash Preview (Non-reasoning)11.6352Sonar Reasoning11.6353Magistral Small 1.211.5354Gemini 2.5 Flash-Lite (Reasoning)11.4355Qwen3 32B (Reasoning)11.4356Ministral 3 14B11.2357
Nanbeige4.1-3B11.2358GPT-4o (Nov '24)11.1359DeepSeek R1 Distill Qwen 32B11.0360Qwen3 VL 32B Instruct11.0361GLM-4.6V (Non-reasoning)10.9362Qwen3 235B A22B (Non-reasoning)10.8363Mistral Small 3.210.7364Gemini 2.0 Flash (experimental)10.6365Magistral Small 110.6366EXAONE 4.0 32B (Reasoning)10.5367Qwen3 VL 8B (Reasoning)10.5368Nova 2.0 Omni (Non-reasoning)10.4369Qwen3 14B (Reasoning)10.4370Llama 4 Scout10.3371DeepSeek R1 0528 Qwen3 8B10.3372Qwen2.5 Max10.1373
Hermes 4 - Llama-3.1 70B (Reasoning)9.9374Gemini 1.5 Pro (Sep '24)9.9375Solar Pro 2 (Preview) (Non-reasoning)9.9376Qwen3 VL 30B A3B Instruct9.9377Gemma 4 E2B (Reasoning)9.8378Claude 3.5 Sonnet (Oct '24)9.8379DeepSeek R1 Distill Llama 70B9.8380
Falcon-H1R-7B9.7381DeepSeek R1 Distill Qwen 14B9.7382GPT-4.1 nano9.6383Ling-flash-2.09.6384Qwen3 Omni 30B A3B (Reasoning)9.5385GPT-4o (Aug '24)9.4386Sonar9.4387Qwen2.5 Instruct 72B9.4388Llama 3.3 Instruct 70B9.3389Step3 VL 10B9.3390Qwen3 30B A3B (Reasoning)9.2391Devstral Small (Jul '25)9.1392Sonar Pro9.1393QwQ 32B-Preview9.1394Ministral 3 8B9.0395Mistral Large 2 (Nov '24)9.0396GLM-4.5V (Reasoning)9.0397Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)8.9398
Granite 4.1 30B8.9399ERNIE 4.5 300B A47B8.9400Qwen3 30B A3B 2507 Instruct8.9401NVIDIA Nemotron 3 Nano 4B8.8402NVIDIA Nemotron Nano 12B v2 VL (Reasoning)8.8403Hermes 4 - Llama-3.1 405B (Reasoning)8.8404Solar Pro 2 (Reasoning)8.8405Gemma 4 E4B (Non-reasoning)8.7406NVIDIA Nemotron Nano 9B V2 (Reasoning)8.7407Hermes 4 - Llama-3.1 405B (Non-reasoning)8.6408Gemini 2.0 Flash-Lite (Feb '25)8.6409Llama Nemotron Super 49B v1.5 (Non-reasoning)8.5410Qwen3 32B (Non-reasoning)8.5411Kimi Linear 48B A3B Instruct8.4412K2-V2 (low)8.4413GPT-4o (May '24)8.4414Gemini 2.0 Flash-Lite (Preview)8.4415Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning)8.4416Llama 3.1 Instruct 405B8.3417Llama 3.3 Nemotron Super 49B v1 (Non-reasoning)8.3418Qwen3 8B (Reasoning)8.3419Qwen3 VL 8B Instruct8.2420Qwen3 4B (Reasoning)8.2421
LFM2.5-8B-A1B8.1422GPT-4o (ChatGPT)8.1423Claude 3.5 Sonnet (June '24)8.1424
Llama 3.1 Tulu3 405B8.1425Ring-flash-2.08.0426Pixtral Large8.0427Olmo 3.1 32B Think7.9428GPT-5 nano (minimal)7.8429Llama 3.1 Instruct 8B7.8430Gemini 1.5 Flash (Sep '24)7.8431Grok 2 (Dec '24)7.8432GPT-4 Turbo7.7433Qwen3 VL 4B (Reasoning)7.7434Solar Pro 2 (Non-reasoning)7.6435Command A7.5436Qwen3.5 2B (Reasoning)7.5437Nova Pro7.5438Llama 3.1 Nemotron Instruct 70B7.4439Gemma 3 27B Instruct7.4440Grok Beta7.3441NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)7.2442NVIDIA Nemotron Nano 9B V2 (Non-reasoning)7.2443Qwen2.5 Instruct 32B7.2444Ministral 3 3B7.1445Mistral Large 2 (Jul '24)7.0446Qwen2.5 Coder Instruct 32B6.9447Qwen3 4B 2507 Instruct6.9448GPT-46.8449GLM-4.5V (Non-reasoning)6.8450Qwen3 14B (Non-reasoning)6.8451Hermes 4 - Llama-3.1 70B (Non-reasoning)6.7452GPT-4o mini6.7453Gemini 2.5 Flash-Lite (Non-reasoning)6.7454Mistral Small 36.7455Nova Lite6.7456Qwen3 30B A3B (Non-reasoning)6.6457Llama 3.1 Instruct 70B6.5458DeepSeek-V2.5 (Dec '24)6.5459Qwen3 4B (Non-reasoning)6.5460Granite 4.1 8B6.4461Sarvam 30B (high)6.4462Gemini 2.0 Flash Thinking Experimental (Dec '24)6.4463DeepSeek-V2.56.4464Gemma 4 E2B (Non-reasoning)6.2465
Phi-4 Mini Instruct6.2466Olmo 3.1 32B Instruct6.2467Mistral Saba6.2468DeepSeek R1 Distill Llama 8B6.2469Gemini 1.5 Pro (May '24)6.1470Olmo 3 32B Think6.1471Llama 3.2 Instruct 90B (Vision)6.0472R1 17766.0473Solar Mini6.0474
Reka Flash (Sep '24)6.0475Qwen2.5 Turbo6.0476Grok-15.8477EXAONE 4.0 32B (Non-reasoning)5.7478Qwen3.5 2B (Non-reasoning)5.7479Qwen2 Instruct 72B5.7480Gemma 3 12B Instruct5.5481Gemini 1.5 Flash-8B5.2482DeepHermes 3 - Mistral 24B Preview (Non-reasoning)5.0483
Jamba 1.7 Large5.0484Qwen3.5 0.8B (Reasoning)5.0485Granite 4.0 H Small4.9486Qwen3 Omni 30B A3B Instruct4.8487Hermes 3 - Llama-3.1 70B4.8488Jamba 1.5 Large4.8489Qwen3 8B (Non-reasoning)4.8490Granite 4.1 3B4.7491DeepSeek-Coder-V24.7492OLMo 2 32B4.7493Jamba 1.6 Large4.7494Phi-44.6495LFM2 24B A2B4.6496Gemini 1.5 Flash (May '24)4.6497Nova Micro4.4498Claude 3 Sonnet4.4499MiniCPM-V 4.6 1.3B4.3500Gemini 1.0 Ultra4.3501Mistral Small (Sep '24)4.3502Phi-3 Mini Instruct 3.8B4.3503Phi-4 Multimodal Instruct4.2504NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)4.2505Gemma 3n E4B Instruct Preview (May '25)4.2506Mistral Large (Feb '24)4.1507Qwen2.5 Coder Instruct 7B 4.1508Mixtral 8x22B Instruct4.0509Llama 3.2 Instruct 3B3.9510Llama 2 Chat 7B3.9511Jamba Reasoning 3B3.8512Reka Flash 33.7513Qwen3 VL 4B Instruct3.7514Qwen1.5 Chat 110B3.7515Olmo 3 7B Think3.6516Qwen3.5 0.8B (Non-reasoning)3.5517Claude 3 Haiku3.5518Claude 2.13.5519OLMo 2 7B3.5520Molmo 7B-D3.4521Ling-mini-2.03.4522Claude 2.03.3523DeepSeek R1 Distill Qwen 1.5B3.3524DeepSeek-V2-Chat3.3525GPT-3.5 Turbo3.2526Mistral Small (Feb '24)3.2527Mistral Medium3.2528Llama 3 Instruct 70B3.1529Llama 3.2 Instruct 11B (Vision)3.0530LFM 40B3.0531
Arctic Instruct3.0532Qwen Chat 72B3.0533PALM-22.8534Gemini 1.0 Pro2.7535DeepSeek Coder V2 Lite Instruct2.7536Llama 2 Chat 70B2.6537Llama 2 Chat 13B2.6538DeepSeek LLM 67B Chat (V1)2.6539
OpenChat 3.5 (1210)2.6540
DBRX Instruct2.6541Sarvam M (Reasoning)2.6542Command-R+ (Apr '24)2.6543Exaone 4.0 1.2B (Reasoning)2.5544Olmo 3 7B Instruct2.4545Exaone 4.0 1.2B (Non-reasoning)2.4546LFM2 2.6B2.3547LFM2.5-1.2B-Instruct2.3548LFM2.5-1.2B-Thinking2.3549Jamba 1.7 Mini2.3550Jamba 1.5 Mini2.3551Granite 4.0 H 1B2.2552Qwen3 1.7B (Reasoning)2.2553Jamba 1.6 Mini2.1554Gemma 3 270M2.0555Granite 4.0 Micro2.0556
Apertus 70B Instruct2.0557Mixtral 8x7B Instruct2.0558DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning)1.9559Llama 65B1.7560Qwen Chat 14B1.7561Claude Instant1.7562Mistral 7B Instruct1.7563Command-R (Mar '24)1.7564Molmo2-8B1.6565Granite 4.0 1B1.6566LFM2 8B A1B1.3567Granite 3.3 8B (Non-reasoning)1.3568Qwen3 1.7B (Non-reasoning)1.1569LFM2.5-VL-1.6B1.0570Granite 4.0 H 350M1.0571Granite 4.0 350M1.0572Apertus 8B Instruct1.0573Tiny Aya Global1.0574Llama 3 Instruct 8B1.0575Llama 3.2 Instruct 1B1.0576Gemma 3 4B Instruct1.0577Gemma 3n E2B Instruct1.0578Gemma 3 1B Instruct1.0579Gemma 3n E4B Instruct1.0580LFM2 1.2B1.0581Qwen3 0.6B (Reasoning)1.0582Qwen3 0.6B (Non-reasoning)1.0
Claude Opus 5 (Adaptive Reasoning, Max Effort)63.12Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)62.53Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)62.14Claude Opus 5 (Adaptive Reasoning, High Effort)61.55GPT-5.6 Sol (max)60.96Kimi K3 (max)59.77GPT-5.6 Sol (xhigh)59.08Claude Opus 5 (Adaptive Reasoning, Medium Effort)58.69Qwen3.8 Max58.110GPT-5.6 Sol (high)57.311Claude Opus 4.8 (Adaptive Reasoning, Max Effort)57.312Muse Spark 1.2 (xhigh)56.813GPT-5.6 Terra (max)56.614GPT-5.5 (xhigh)56.315Grok 4.5 (high)55.816GPT-5.6 Sol (medium)55.617Claude Sonnet 5 (Adaptive Reasoning, Max Effort)55.318Claude Opus 4.7 (Adaptive Reasoning, Max Effort)55.019GPT-5.5 (high)54.720Muse Spark 1.1 (xhigh)53.221GPT-5.4 (xhigh)53.122GPT-5.6 Terra (xhigh)52.823GLM-5.2 (max)52.624Claude Opus 5 (Adaptive Reasoning, Low Effort)52.525GPT-5.6 Luna (max)52.326Gemini 3.5 Flash (high)52.027DeepSeek V4 Flash 0731 (Reasoning, Max Effort)51.828Gemini 3.6 Flash (high)51.629GPT-5.5 (medium)51.430GPT-5.6 Sol (low)50.731GPT-5.6 Terra (high)50.132GPT-5.6 Luna (xhigh)50.133Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)48.434Kimi K3 (low)48.335Gemini 3.1 Pro Preview47.736GPT-5.6 Luna (high)47.037GPT-5.6 Terra (medium)46.838Gemini 3.5 Flash (medium)46.739Qwen3.7 Max46.740GPT-5.3 Codex (xhigh)45.541MiniMax-M345.442DeepSeek V4 Pro (Reasoning, Max Effort)45.343Kimi K2.645.144Motif 3 (Beta)44.945Claude Opus 4.6 (Adaptive Reasoning, Max Effort)44.946GPT-5.5 (low)44.547Claude Opus 4.7 (Non-reasoning, High Effort)43.948DeepSeek V4 Pro (Reasoning, High Effort)43.749Muse Spark43.550GPT-5.2 (xhigh)43.351Kimi K2.7 Code43.052MiMo-V2.5-Pro42.953Claude Sonnet 5 (Non-reasoning, High Effort)42.654Inkling (xhigh)42.355Hy342.256DeepSeek V4 Flash (Reasoning, Max Effort)42.157GPT-5.6 Sol (Non-reasoning)41.958Claude Opus 4.5 (Reasoning)41.959Nex-N2-Pro41.760MiMo-V2-Pro41.461GPT-5.6 Terra (low)41.362Inkling Small41.263GPT-5.2 Codex (xhigh)41.264Qwen3.6 Max Preview41.165GLM-5.1 (Reasoning)41.066GPT-5.4 mini (xhigh)40.967Grok Build 0.1 061640.768Gemini 3 Pro Preview (high)40.669GLM-5 (Reasoning)40.670Qwen3.6 Plus40.571GPT-5.4 (low)40.272JT-4.1 Flash 236B A21B39.873Agnes 2.5 Pro Alpha39.774GPT-5.4 nano (xhigh)39.775Qwen3.7 Plus39.476GLM-5-Turbo39.177DeepSeek V4 Flash (Reasoning, High Effort)39.078GPT-5.6 Luna (medium)38.979GPT-5.2 (medium)38.980MiniMax-M2.738.981Claude Opus 4.6 (Non-reasoning, High Effort)38.882Gemini 3 Flash Preview (Reasoning)38.783Nemotron 3 Ultra 550B A55B (Reasoning)38.384MiMo-V2.538.085Grok 4.20 0309 v2 (Reasoning)38.086Grok 4.3 (high)37.987Ling-3.0-flash37.888Qwen3.6 27B (Reasoning)37.789GPT-5.1 (high)37.590Gemini 3.5 Flash-Lite37.491Claude 4.5 Sonnet (Reasoning)37.492Grok 4.20 0309 (Reasoning)37.493MiMo-V2-Omni-032737.394GPT-5 Codex (high)37.095Grok 4.3 (medium)36.996Claude Sonnet 4.6 (Non-reasoning, High Effort)36.897Grok 4.3 (low)36.398GLM-5.1 (Non-reasoning)36.399Kimi K2.5 (Reasoning)36.0100MiMo-V2-Omni35.9101Gemini 3.5 Flash (minimal)35.8102GPT-5.5 (Non-reasoning)35.8103GPT-5.1 Codex (high)35.6104Claude Opus 4.5 (Non-reasoning)35.6105Kimi K2.6 (Non-reasoning)35.4106GPT-5 (high)35.3107GLM 5V Turbo (Reasoning)35.3108Claude Sonnet 4.6 (Non-reasoning, Low Effort)35.1109GLM-5.2 (Non-reasoning)34.8110GPT-5.6 Terra (Non-reasoning)34.6111GPT-5 (medium)34.6112Qwen3.5 27B (Reasoning)34.6113Claude 4.1 Opus (Reasoning)34.5114MiniMax-M2.534.5115GLM-4.7 (Reasoning)34.5116Hy3-preview (Reasoning)34.4117Qwen3.5 397B A17B (Reasoning)34.3118GPT-5.5 Instant (May 2026)34.3119Grok 434.1120MiMo-V2-Flash (Feb 2026)34.0121GPT-5.6 Luna (low)33.9122KAT Coder Pro V233.9123LongCat 2.033.9124Gemini 3 Pro Preview (low)33.9125Kimi K2 Thinking33.5126o3-pro33.3127GLM-5 (Non-reasoning)33.2128Qwen3.5 122B A10B (Reasoning)32.8129Qwen3.5 397B A17B (Non-reasoning)32.7130DeepSeek V3.2 (Reasoning)32.6131Qwen3 Max Thinking32.5132Qwen3.6 35B A3B (Reasoning)32.1133MiniMax-M2.132.1134DeepSeek V4 Pro (Non-reasoning)31.9135GPT-5 (low)31.9136MiMo-V2-Flash (Reasoning)31.9137Claude 4 Opus (Reasoning)31.7138GPT-5 mini (medium)31.6139G9v3-39A5B31.4140Qwen3.6 27B (Non-reasoning)31.3141Qwen3.5 Omni Plus31.3142GPT-5.1 Codex mini (high)31.3143Grok 4.1 Fast (Reasoning)31.3144o331.1145Ring-2.6-1T31.1146Step 3.7 Flash30.9147GPT-5.4 nano (medium)30.8148DeepSeek V3.1 Terminus (Reasoning)30.6149GPT-5.4 mini (medium)30.5150Mistral Medium 3.530.4151Kimi K2.5 (Non-reasoning)30.1152Qwen3.5 27B (Non-reasoning)30.0153Claude 4.5 Haiku (Reasoning)29.9154Claude 4.5 Sonnet (Non-reasoning)29.9155Qwen3.5 35B A3B (Reasoning)29.9156Claude 4 Sonnet (Reasoning)29.8157Gemma 4 31B (Reasoning)29.7158DeepSeek V4 Flash (Non-reasoning)29.3159GLM-4.6 (Reasoning)29.3160GPT-5.5 Instant (June 2026)29.2161JT-35B-Flash29.0162KAT-Coder-Pro V128.9163MiniMax-M228.9164Claude 4.1 Opus (Non-reasoning)28.8165MiMo-V2.5-Pro (Non-reasoning)28.4166GPT-5.4 (Non-reasoning)28.3167Qwen3.5 122B A10B (Non-reasoning)28.2168Gemini 3 Flash Preview (Non-reasoning)27.9169Grok 4 Fast (Reasoning)27.9170Claude 3.7 Sonnet (Reasoning)27.6171GLM-4.7 (Non-reasoning)27.1172GPT-5.6 Luna (Non-reasoning)26.8173Hy3-preview (Non-reasoning)26.6174Ling-2.6-1T26.6175Doubao Seed Code26.5176GPT-5.2 (Non-reasoning)26.5177Step 3.5 Flash 260326.5178Gemma 4 26B A4B (Reasoning)26.1179o4-mini (high)26.1180Claude 4 Opus (Non-reasoning)26.0181Claude 4 Sonnet (Non-reasoning)26.0182Step 3.5 Flash26.0183Gemini 2.5 Pro25.9184DeepSeek V3.2 Exp (Reasoning)25.9185GPT-5 mini (high)25.8186NVIDIA Nemotron 3 Super 120B A12B (Reasoning)25.7187Gemini 3.1 Flash-Lite25.6188Qwen3 Max Thinking (Preview)25.5189DeepSeek V3.2 (Non-reasoning)25.1190Grok 4.3 (Non-reasoning)25.0191MiMo-V2-Flash (Non-reasoning)25.0192Qwen3.6 35B A3B (Non-reasoning)24.5193Qwen3 Max24.5194Qwen3.5 35B A3B (Non-reasoning)24.3195Ling 3.0 Tiny24.3196Gemini 2.5 Flash Preview (Sep '25) (Reasoning)24.2197gpt-oss-120b (high)24.1198Claude 4.5 Haiku (Non-reasoning)24.1199Kimi K2 090524.0200o123.9201Claude 3.7 Sonnet (Non-reasoning)23.9202Gemini 2.5 Pro Preview (Mar' 25)23.4203GLM-4.6 (Non-reasoning)23.4204GLM-4.7-Flash (Reasoning)23.3205Grok 3 mini Reasoning (high)22.9206Grok 4.20 0309 (Non-reasoning)22.9207Command A+22.8208Gemini 2.5 Pro Preview (May' 25)22.7209DeepSeek V3.2 Speciale22.6210K-EXAONE (Reasoning)22.4211Gemma 4 31B (Non-reasoning)22.3212Gemma 4 12B (Reasoning)22.3213ERNIE 5.0 Thinking Preview22.3214Grok 4.20 0309 v2 (Non-reasoning)22.2215Nova 2.0 Pro Preview (medium)22.0216Grok Code Fast 122.0217Mercury 221.9218Qwen3.5 9B (Reasoning)21.8219DeepSeek V3.1 Terminus (Non-reasoning)21.7220DeepSeek V3.2 Exp (Non-reasoning)21.7221Apriel-v1.5-15B-Thinker21.6222DeepSeek V3.1 (Non-reasoning)21.4223Nova 2.0 Omni (medium)21.3224Qwen3 Coder Next21.3225DeepSeek V3.1 (Reasoning)21.0226Qwen3 VL 235B A22B (Reasoning)20.9227Apriel-v1.6-15B-Thinker20.8228GPT-5.1 (Non-reasoning)20.7229Qwen3.5 9B (Non-reasoning)20.6230EXAONE 4.5 33B20.5231Gemma 4 26B A4B (Non-reasoning)20.4232Qwen3.5 4B (Reasoning)20.4233DeepSeek R1 0528 (May '25)20.4234Gemini 2.5 Flash (Reasoning)20.3235North Mini Code20.2236GPT-5 nano (high)20.1237Nova 2.0 Pro Preview (low)20.0238Qwen3 235B A22B 2507 (Reasoning)19.9239Mistral Small 4 (Reasoning)19.7240Kimi K219.7241GLM-4.5 (Reasoning)19.7242GPT-4.119.6243Qwen3 Max (Preview)19.4244Devstral 219.2245Nova 2.0 Lite (medium)19.2246Qwen3.5 Omni Flash19.2247GPT-5 nano (medium)19.2248o3-mini19.2249o1-pro19.1250Gemini 2.5 Flash Preview (Sep '25) (Non-reasoning)19.1251JT-MINI18.8252Trinity Large Thinking18.6253DeepSeek R1 (Jan '25)18.6254Grok 318.6255Seed-OSS-36B-Instruct18.5256Nova 2.0 Lite (high)18.4257Qwen3 235B A22B 2507 Instruct18.4258HyperNova 60B 260518.3259Qwen3 Coder 480B A35B Instruct18.2260Qwen3 VL 32B (Reasoning)18.1261Magistral Medium 1.218.0262Nova 2.0 Lite (low)18.0263Nemotron Cascade 2 30B A3B18.0264Sonar Reasoning Pro18.0265MiniMax M1 80k17.9266GPT-5.4 nano (Non-Reasoning)17.8267Devstral Small 217.7268Gemini 2.5 Flash Preview (Reasoning)17.7269LongCat Flash Lite17.4270GPT-5 (minimal)17.3271K2 Think V217.2272HyperCLOVA X SEED Think (32B)17.2273o1-preview17.2274Grok 4.1 Fast (Non-reasoning)17.0275K-EXAONE (Non-reasoning)16.9276Qwen3 Next 80B A3B (Reasoning)16.9277GLM-4.6V (Reasoning)16.9278GPT-5.4 mini (Non-Reasoning)16.8279Nova 2.0 Omni (low)16.7280GLM-4.5-Air16.7281Mi:dm K 2.5 Pro16.6282Grok 4 Fast (Non-reasoning)16.6283G9v3-3B16.5284Ring-1T16.3285Qwen3.5 4B (Non-reasoning)16.1286Mistral Large 315.9287INTELLECT-315.7288o3-mini (high)15.7289GLM-4.7-Flash (Non-reasoning)15.6290GPT-5 (ChatGPT)15.4291gpt-oss-20b (high)15.2292Solar Open 100B (Reasoning)15.2293Gemini 2.5 Flash-Lite Preview (Sep '25) (Reasoning)15.2294DeepSeek V3 032415.2295Grok 3 Reasoning Beta15.2296Nemotron 3 Nano Omni 30B A3B Reasoning15.0297gpt-oss-120b (low)14.9298Mistral Small 3.114.9299GPT-4.1 mini14.8300Mistral Medium 3.114.7301Qwen3 30B A3B 2507 (Reasoning)14.6302Llama 4 Maverick14.5303Solar Pro 314.5304NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)14.5305MiniMax M1 40k14.5306gpt-oss-20b (low)14.4307Nova 2.0 Pro Preview (Non-reasoning)14.4308Qwen3 VL 235B A22B Instruct14.4309GPT-5 mini (minimal)14.3310K2-V2 (high)14.2311Ling 2.6 Flash14.2312Gemini 2.5 Flash (Non-reasoning)14.2313DeepSeek V3 (Dec '24)14.2314o1-mini14.0315DiffusionGemma 26B A4B13.8316Qwen3 Next 80B A3B Instruct13.8317GPT-4.5 (Preview)13.6318Tri-21B-think Preview13.6319Qwen3 Coder 30B A3B Instruct13.6320Qwen3 235B A22B (Reasoning)13.5321QwQ 32B13.4322Qwen3 VL 30B A3B (Reasoning)13.4323Gemini 2.0 Flash Thinking Experimental (Jan '25)13.3324Gemma 4 12B (Non-reasoning)13.2325Gemini 2.5 Flash-Lite Preview (Sep '25) (Non-reasoning)13.1326Motif-2-12.7B-Reasoning12.8327Nova Premier12.7328Ling-1T12.7329Celeris-112.5330Magistral Medium 112.5331Mistral Medium 312.5332Solar Pro 2 (Preview) (Reasoning)12.5333Gemma 4 E4B (Reasoning)12.4334Llama Nemotron Super 49B v1.5 (Reasoning)12.4335K2-V2 (medium)12.4336Claude 3.5 Haiku12.4337Devstral Medium12.4338Mistral Small 4 (Non-reasoning)12.3339Tri-21B-Think12.3340GPT-4o (March 2025, chatgpt-4o-latest)12.3341Gemini 2.0 Flash (Feb '25)12.2342Llama 3.3 Nemotron Super 49B v1 (Reasoning)12.2343Sarvam 105B (high)11.9344MiniCPM5-1B (Reasoning)11.9345Qwen3 4B 2507 (Reasoning)11.9346Nova 2.0 Lite (Non-reasoning)11.8347Gemini 2.0 Pro Experimental (Feb '25)11.8348Claude 3 Opus11.8349Devstral Small (May '25)11.8350MiniCPM5-1B (Non-reasoning)11.7351Gemini 2.5 Flash Preview (Non-reasoning)11.6352Sonar Reasoning11.6353Magistral Small 1.211.5354Gemini 2.5 Flash-Lite (Reasoning)11.4355Qwen3 32B (Reasoning)11.4356Ministral 3 14B11.2357Nanbeige4.1-3B11.2358GPT-4o (Nov '24)11.1359DeepSeek R1 Distill Qwen 32B11.0360Qwen3 VL 32B Instruct11.0361GLM-4.6V (Non-reasoning)10.9362Qwen3 235B A22B (Non-reasoning)10.8363Mistral Small 3.210.7364Gemini 2.0 Flash (experimental)10.6365Magistral Small 110.6366EXAONE 4.0 32B (Reasoning)10.5367Qwen3 VL 8B (Reasoning)10.5368Nova 2.0 Omni (Non-reasoning)10.4369Qwen3 14B (Reasoning)10.4370Llama 4 Scout10.3371DeepSeek R1 0528 Qwen3 8B10.3372Qwen2.5 Max10.1373Hermes 4 - Llama-3.1 70B (Reasoning)9.9374Gemini 1.5 Pro (Sep '24)9.9375Solar Pro 2 (Preview) (Non-reasoning)9.9376Qwen3 VL 30B A3B Instruct9.9377Gemma 4 E2B (Reasoning)9.8378Claude 3.5 Sonnet (Oct '24)9.8379DeepSeek R1 Distill Llama 70B9.8380Falcon-H1R-7B9.7381DeepSeek R1 Distill Qwen 14B9.7382GPT-4.1 nano9.6383Ling-flash-2.09.6384Qwen3 Omni 30B A3B (Reasoning)9.5385GPT-4o (Aug '24)9.4386Sonar9.4387Qwen2.5 Instruct 72B9.4388Llama 3.3 Instruct 70B9.3389Step3 VL 10B9.3390Qwen3 30B A3B (Reasoning)9.2391Devstral Small (Jul '25)9.1392Sonar Pro9.1393QwQ 32B-Preview9.1394Ministral 3 8B9.0395Mistral Large 2 (Nov '24)9.0396GLM-4.5V (Reasoning)9.0397Llama 3.1 Nemotron Ultra 253B v1 (Reasoning)8.9398Granite 4.1 30B8.9399ERNIE 4.5 300B A47B8.9400Qwen3 30B A3B 2507 Instruct8.9401NVIDIA Nemotron 3 Nano 4B8.8402NVIDIA Nemotron Nano 12B v2 VL (Reasoning)8.8403Hermes 4 - Llama-3.1 405B (Reasoning)8.8404Solar Pro 2 (Reasoning)8.8405Gemma 4 E4B (Non-reasoning)8.7406NVIDIA Nemotron Nano 9B V2 (Reasoning)8.7407Hermes 4 - Llama-3.1 405B (Non-reasoning)8.6408Gemini 2.0 Flash-Lite (Feb '25)8.6409Llama Nemotron Super 49B v1.5 (Non-reasoning)8.5410Qwen3 32B (Non-reasoning)8.5411Kimi Linear 48B A3B Instruct8.4412K2-V2 (low)8.4413GPT-4o (May '24)8.4414Gemini 2.0 Flash-Lite (Preview)8.4415Llama 3.1 Nemotron Nano 4B v1.1 (Reasoning)8.4416Llama 3.1 Instruct 405B8.3417Llama 3.3 Nemotron Super 49B v1 (Non-reasoning)8.3418Qwen3 8B (Reasoning)8.3419Qwen3 VL 8B Instruct8.2420Qwen3 4B (Reasoning)8.2421LFM2.5-8B-A1B8.1422GPT-4o (ChatGPT)8.1423Claude 3.5 Sonnet (June '24)8.1424Llama 3.1 Tulu3 405B8.1425Ring-flash-2.08.0426Pixtral Large8.0427Olmo 3.1 32B Think7.9428GPT-5 nano (minimal)7.8429Llama 3.1 Instruct 8B7.8430Gemini 1.5 Flash (Sep '24)7.8431Grok 2 (Dec '24)7.8432GPT-4 Turbo7.7433Qwen3 VL 4B (Reasoning)7.7434Solar Pro 2 (Non-reasoning)7.6435Command A7.5436Qwen3.5 2B (Reasoning)7.5437Nova Pro7.5438Llama 3.1 Nemotron Instruct 70B7.4439Gemma 3 27B Instruct7.4440Grok Beta7.3441NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)7.2442NVIDIA Nemotron Nano 9B V2 (Non-reasoning)7.2443Qwen2.5 Instruct 32B7.2444Ministral 3 3B7.1445Mistral Large 2 (Jul '24)7.0446Qwen2.5 Coder Instruct 32B6.9447Qwen3 4B 2507 Instruct6.9448GPT-46.8449GLM-4.5V (Non-reasoning)6.8450Qwen3 14B (Non-reasoning)6.8451Hermes 4 - Llama-3.1 70B (Non-reasoning)6.7452GPT-4o mini6.7453Gemini 2.5 Flash-Lite (Non-reasoning)6.7454Mistral Small 36.7455Nova Lite6.7456Qwen3 30B A3B (Non-reasoning)6.6457Llama 3.1 Instruct 70B6.5458DeepSeek-V2.5 (Dec '24)6.5459Qwen3 4B (Non-reasoning)6.5460Granite 4.1 8B6.4461Sarvam 30B (high)6.4462Gemini 2.0 Flash Thinking Experimental (Dec '24)6.4463DeepSeek-V2.56.4464Gemma 4 E2B (Non-reasoning)6.2465Phi-4 Mini Instruct6.2466Olmo 3.1 32B Instruct6.2467Mistral Saba6.2468DeepSeek R1 Distill Llama 8B6.2469Gemini 1.5 Pro (May '24)6.1470Olmo 3 32B Think6.1471Llama 3.2 Instruct 90B (Vision)6.0472R1 17766.0473Solar Mini6.0474Reka Flash (Sep '24)6.0475Qwen2.5 Turbo6.0476Grok-15.8477EXAONE 4.0 32B (Non-reasoning)5.7478Qwen3.5 2B (Non-reasoning)5.7479Qwen2 Instruct 72B5.7480Gemma 3 12B Instruct5.5481Gemini 1.5 Flash-8B5.2482DeepHermes 3 - Mistral 24B Preview (Non-reasoning)5.0483Jamba 1.7 Large5.0484Qwen3.5 0.8B (Reasoning)5.0485Granite 4.0 H Small4.9486Qwen3 Omni 30B A3B Instruct4.8487Hermes 3 - Llama-3.1 70B4.8488Jamba 1.5 Large4.8489Qwen3 8B (Non-reasoning)4.8490Granite 4.1 3B4.7491DeepSeek-Coder-V24.7492OLMo 2 32B4.7493Jamba 1.6 Large4.7494Phi-44.6495LFM2 24B A2B4.6496Gemini 1.5 Flash (May '24)4.6497Nova Micro4.4498Claude 3 Sonnet4.4499MiniCPM-V 4.6 1.3B4.3500Gemini 1.0 Ultra4.3501Mistral Small (Sep '24)4.3502Phi-3 Mini Instruct 3.8B4.3503Phi-4 Multimodal Instruct4.2504NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)4.2505Gemma 3n E4B Instruct Preview (May '25)4.2506Mistral Large (Feb '24)4.1507Qwen2.5 Coder Instruct 7B 4.1508Mixtral 8x22B Instruct4.0509Llama 3.2 Instruct 3B3.9510Llama 2 Chat 7B3.9511Jamba Reasoning 3B3.8512Reka Flash 33.7513Qwen3 VL 4B Instruct3.7514Qwen1.5 Chat 110B3.7515Olmo 3 7B Think3.6516Qwen3.5 0.8B (Non-reasoning)3.5517Claude 3 Haiku3.5518Claude 2.13.5519OLMo 2 7B3.5520Molmo 7B-D3.4521Ling-mini-2.03.4522Claude 2.03.3523DeepSeek R1 Distill Qwen 1.5B3.3524DeepSeek-V2-Chat3.3525GPT-3.5 Turbo3.2526Mistral Small (Feb '24)3.2527Mistral Medium3.2528Llama 3 Instruct 70B3.1529Llama 3.2 Instruct 11B (Vision)3.0530LFM 40B3.0531Arctic Instruct3.0532Qwen Chat 72B3.0533PALM-22.8534Gemini 1.0 Pro2.7535DeepSeek Coder V2 Lite Instruct2.7536Llama 2 Chat 70B2.6537Llama 2 Chat 13B2.6538DeepSeek LLM 67B Chat (V1)2.6539OpenChat 3.5 (1210)2.6540DBRX Instruct2.6541Sarvam M (Reasoning)2.6542Command-R+ (Apr '24)2.6543Exaone 4.0 1.2B (Reasoning)2.5544Olmo 3 7B Instruct2.4545Exaone 4.0 1.2B (Non-reasoning)2.4546LFM2 2.6B2.3547LFM2.5-1.2B-Instruct2.3548LFM2.5-1.2B-Thinking2.3549Jamba 1.7 Mini2.3550Jamba 1.5 Mini2.3551Granite 4.0 H 1B2.2552Qwen3 1.7B (Reasoning)2.2553Jamba 1.6 Mini2.1554Gemma 3 270M2.0555Granite 4.0 Micro2.0556Apertus 70B Instruct2.0557Mixtral 8x7B Instruct2.0558DeepHermes 3 - Llama-3.1 8B Preview (Non-reasoning)1.9559Llama 65B1.7560Qwen Chat 14B1.7561Claude Instant1.7562Mistral 7B Instruct1.7563Command-R (Mar '24)1.7564Molmo2-8B1.6565Granite 4.0 1B1.6566LFM2 8B A1B1.3567Granite 3.3 8B (Non-reasoning)1.3568Qwen3 1.7B (Non-reasoning)1.1569LFM2.5-VL-1.6B1.0570Granite 4.0 H 350M1.0571Granite 4.0 350M1.0572Apertus 8B Instruct1.0573Tiny Aya Global1.0574Llama 3 Instruct 8B1.0575Llama 3.2 Instruct 1B1.0576Gemma 3 4B Instruct1.0577Gemma 3n E2B Instruct1.0578Gemma 3 1B Instruct1.0579Gemma 3n E4B Instruct1.0580LFM2 1.2B1.0581Qwen3 0.6B (Reasoning)1.0582Qwen3 0.6B (Non-reasoning)1.0数值越大越好 · 可上下滚动查看更多
智能指数 vs 价格
性价比横轴：综合价格（$/1M，对数坐标）· 纵轴：智能指数 · 滚轮/双指捏合以屏幕中心等比缩放（带缓动）
Alibaba (15)Google (9)Anthropic (6)OpenAI (6)SpaceXAI (5)InclusionAI (4)Kimi (3)Z AI (3)其他 (29)趋势线（R² 0.62）
悬停查看详情，点击跳转模型页 · 手机双指捏合缩放、单指拖动平移，桌面滚轮缩放 · 虚线为线性拟合趋势线
默认按发布日期倒序排序，点击任意列表头可切换排序 · 同一型号只保留最强等级
| # | 模型 | 供应商 | 发布日期↓ | 价格-输入 | 价格-输出 | 输出速度 | 首token时间 | 首答案token时间 | 智能指数 | 代码指数 | 数学指数 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Ling 3.0 Tiny | InclusionAI | 2026-08-06 | - | - | - | - | - | 24.30 | 26.50 | - |
| 2 | Muse Spark 1.2 (xhigh) | Meta | 2026-08-05 | ¥8.45 | ¥28.73 | - | - | - | 56.80 | 72.20 | - |
| 3 | Ling-3.0-flash | InclusionAI | 2026-08-04 | ¥0.51 | ¥1.49 | 381 | 1.62s | 6.88s | 37.80 | 50.60 | - |
| 4 | G9v3-39A5B | AI9Stars | 2026-08-03 | - | - | - | - | - | 31.40 | 31.70 | - |
| 5 | Qwen3.8 Max | Alibaba | 2026-08-03 | ¥13.52 | ¥40.56 | 70 | 1.93s | 30.66s | 58.10 | 71.80 | - |
| 6 | DeepSeek V4 Flash 0731 (Reasoning, Max Effort) | DeepSeek | 2026-07-31 | ¥0.95 | ¥1.89 | 106 | 0.97s | 19.79s | 51.80 | 69.10 | - |
| 7 | Inkling Small | Thinking Machines | 2026-07-30 | ¥2.03 | ¥8.11 | 119 | 1.45s | 18.25s | 41.20 | 52.90 | - |
| 8 | Claude Opus 5 (Adaptive Reasoning, Max Effort) | Anthropic | 2026-07-24 | ¥33.80 | ¥168.99 | 53 | 40.42s | 40.42s | 63.10 | 78.00 | - |
| 9 | Agnes 2.5 Pro Alpha | Sapiens AI | 2026-07-24 | ¥3.04 | ¥6.08 | 131 | 3.46s | 18.69s | 39.70 | 58.80 | - |
| 10 | Celeris-1 | Celeris | 2026-07-24 | ¥1.35 | ¥4.73 | 2,037 | 0.48s | 0.48s | 12.50 | 14.40 | - |
| 11 | G9v3-3B | AI9Stars | 2026-07-23 | - | - | - | - | - | 16.50 | 9.90 | - |
| 12 | Gemini 3.6 Flash (high) | Google | 2026-07-21 | ¥10.14 | ¥50.70 | 203 | 13.56s | 13.56s | 51.60 | 69.20 | - |
| 13 | Gemini 3.5 Flash-Lite | Google | 2026-07-21 | ¥2.03 | ¥16.90 | 360 | 8.09s | 8.09s | 37.40 | 49.30 | - |
| 14 | Kimi K3 (max) | Kimi | 2026-07-16 | ¥20.28 | ¥101.40 | 37 | 2.45s | 56.05s | 59.70 | 76.20 | - |
| 15 | Inkling (xhigh) | Thinking Machines | 2026-07-15 | ¥6.76 | ¥27.38 | 81 | 2.18s | 26.97s | 42.30 | 52.10 | - |
| 16 | Motif 3 (Beta) | Motif Technologies | 2026-07-14 | - | - | - | - | - | 44.90 | 62.00 | - |
| 17 | GPT-5.6 Terra (max) | OpenAI | 2026-07-09 | ¥13.52 | ¥81.12 | 125 | 88.68s | 88.68s | 56.60 | 76.70 | - |
| 18 | GPT-5.6 Luna (max) | OpenAI | 2026-07-09 | ¥1.35 | ¥8.11 | 177 | 69.79s | 69.79s | 52.30 | 71.40 | - |
| 19 | GPT-5.6 Sol (max) | OpenAI | 2026-07-09 | ¥33.80 | ¥202.79 | 63 | 64.31s | 64.31s | 60.90 | 77.40 | - |
| 20 | Muse Spark 1.1 (xhigh) | Meta | 2026-07-09 | ¥8.45 | ¥28.73 | 199 | 1.43s | 11.48s | 53.20 | 71.30 | - |
| 21 | JT-4.1 Flash 236B A21B | China Mobile | 2026-07-09 | - | - | - | - | - | 39.80 | 52.40 | - |
| 22 | Grok 4.5 (high) | SpaceXAI | 2026-07-08 | ¥13.52 | ¥40.56 | 54 | 12.43s | 12.43s | 55.80 | 72.40 | - |
| 23 | Hy3 | Tencent | 2026-07-06 | ¥0.92 | ¥3.77 | 67 | 1.73s | 31.71s | 42.20 | 58.80 | - |
| 24 | Claude Sonnet 5 (Adaptive Reasoning, Max Effort) | Anthropic | 2026-06-30 | ¥13.52 | ¥67.60 | 75 | 137.72s | 137.72s | 55.30 | 71.50 | - |
| 25 | LongCat 2.0 | LongCat | 2026-06-29 | ¥5.07 | ¥19.94 | 43 | 1.99s | 48.28s | 33.90 | 45.30 | - |
| 26 | GLM-5.2 (max) | Z AI | 2026-06-16 | ¥9.13 | ¥29.00 | 119 | 1.07s | 17.87s | 52.60 | 68.80 | - |
| 27 | Grok Build 0.1 0616 | SpaceXAI | 2026-06-16 | ¥6.76 | ¥13.52 | - | - | - | 40.70 | 51.50 | - |
| 28 | Kimi K2.7 Code | Kimi | 2026-06-12 | ¥6.42 | ¥27.04 | 38 | 1.49s | 59.64s | 43.00 | 60.80 | - |
| 29 | DiffusionGemma 26B A4B | Google | 2026-06-10 | - | - | - | - | - | 13.80 | 19.70 | - |
| 30 | Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) | Anthropic | 2026-06-09 | ¥67.60 | ¥337.98 | 64 | 78.69s | 78.69s | 62.10 | 76.50 | - |
| 31 | North Mini Code | Cohere | 2026-06-09 | - | - | 24 | 0.41s | 82.87s | 20.20 | 36.50 | - |
| 32 | Nemotron 3 Ultra 550B A55B (Reasoning) | NVIDIA | 2026-06-04 | ¥4.06 | ¥18.59 | 136 | 1.50s | 18.29s | 38.30 | 49.30 | - |
| 33 | Gemma 4 12B (Reasoning) | Google | 2026-06-03 | ¥0.68 | ¥2.03 | 44 | 1.53s | 46.84s | 22.30 | 31.00 | - |
| 34 | Nex-N2-Pro | Nex AGI | 2026-06-02 | ¥3.38 | ¥16.90 | 117 | 0.86s | 17.95s | 41.70 | 59.10 | - |
| 35 | MiniMax-M3 | MiniMax | 2026-06-01 | ¥2.03 | ¥8.11 | 81 | 1.16s | 25.95s | 45.40 | 58.60 | - |
| 36 | Qwen3.7 Plus | Alibaba | 2026-06-01 | ¥2.70 | ¥10.82 | 55 | 1.38s | 37.50s | 39.40 | 55.90 | - |
| 37 | Step 3.7 Flash | StepFun | 2026-05-29 | ¥1.35 | ¥7.77 | 386 | 0.75s | 5.93s | 30.90 | 39.60 | - |
| 38 | LFM2.5-8B-A1B | Liquid AI | 2026-05-28 | - | - | 339 | 1.70s | 7.59s | 8.10 | - | - |
| 39 | Claude Opus 4.8 (Adaptive Reasoning, Max Effort) | Anthropic | 2026-05-28 | ¥33.80 | ¥168.99 | - | - | - | 57.30 | 74.30 | - |
| 40 | HyperNova 60B 2605 | Multiverse Computing | 2026-05-26 | ¥0.27 | ¥0.95 | 416 | 0.51s | 5.32s | 18.30 | 23.20 | - |
共显示 40 / 403 个模型 · 数据来源：Artificial Analysis 公开评测