Fastest AI models: compare output speed
Models ranked by reported output tokens per second. Output speed is different from time to first answer and total response time, especially for reasoning models. Missing measurements are excluded.
Which should you choose?
No single benchmark covers every use case. Try your own prompts and compare accuracy, response time, input/output support, and the actual bill.
AI models
| Model | Intelligence score | Benchmark cost (USD) | Input price / 1M tokens (USD) | Output price / 1M tokens (USD) | Output speed |
|---|---|---|---|---|---|
| Celeris-1Celeris | 6.30 | $60.0 | 0.20 | 0.70 | 1252 tok/s |
| Mercury 2Inception | 11.50 | $153 | 0.25 | 0.75 | 925 tok/s |
| Gemini 2.5 Flash-Lite (Reasoning)Google | 8.50 | - | 0.10 | 0.40 | 401 tok/s |
| Gemini 3.5 Flash-LiteGoogle | 22.70 | $267 | 0.30 | 2.50 | 363 tok/s |
| HyperNova 60B 2605 (high, based on gpt-oss-120b)Multiverse Computing | 11.70 | - | 0.04 | 0.14 | 353 tok/s |
| Gemini 3.7 Flash (high)Google | 39.40 | $1084 | 0.75 | 3.75 | 322 tok/s |
| Trinity Large ThinkingArcee AI | 10.90 | $518 | 0.25 | 0.90 | 320 tok/s |
| Gemini 3.7 Flash (low)Google | 36.90 | - | 0.75 | 3.75 | 320 tok/s |
| Gemini 2.5 Flash-Lite (Non-reasoning)Google | 6.70 | - | 0.10 | 0.40 | 317 tok/s |
| Gemini 3.1 Flash-LiteGoogle | 16.00 | $119 | 0.25 | 1.50 | 315 tok/s |
| Gemini 3.7 Flash (medium)Google | 39.60 | - | 0.75 | 3.75 | 311 tok/s |
| Ling 3.0 FlashInclusionAI | 24.90 | - | 0.07 | 0.22 | 310 tok/s |
| Nemotron 3.5 LightningNVIDIA | 13.60 | $105 | 0.06 | 0.20 | 289 tok/s |
| Nova MicroAmazon | 5.90 | - | 0.04 | 0.14 | 286 tok/s |
| Gemini 3.8 Flash (high)Google | 41.20 | $1623 | 0.75 | 3.75 | 281 tok/s |
| Gemini 3.8 Flash (low)Google | 33.80 | - | 0.75 | 3.75 | 265 tok/s |
| gpt-oss-20b (low)OpenAI | 10.00 | - | 0.07 | 0.20 | 249 tok/s |
| Qwen3.5 Omni FlashAlibaba | 12.50 | - | 0.10 | 0.80 | 248 tok/s |
| gpt-oss-120b (low)OpenAI | 10.20 | - | 0.15 | 0.54 | 247 tok/s |
| Gemini 3.8 Flash (medium)Google | 40.00 | $1100 | 0.75 | 3.75 | 246 tok/s |
| Muse Spark 1.2 (xhigh)Meta | 39.80 | $1385 | 1.25 | 4.25 | 238 tok/s |
| Muse Spark 1.3 (max)MetaBest value frontier | 48.20 | $2000 | 1.25 | 4.25 | 229 tok/s |
| Gemini 2.5 Flash (Reasoning)Google | 13.10 | - | 0.30 | 2.50 | 229 tok/s |
| NVIDIA Nemotron 3 Nano 30B A3B (Non-reasoning)NVIDIA | 6.80 | - | 0.05 | 0.20 | 227 tok/s |
| Step 3.5 FlashStepFun | 16.60 | - | 0.10 | 0.30 | 226 tok/s |
| Gemini 3.5 Flash (medium)Google | 33.60 | - | 1.50 | 9.00 | 219 tok/s |
| GPT-5.4 mini (xhigh)OpenAI | 24.60 | $1438 | 0.75 | 4.50 | 218 tok/s |
| Step 3.5 Flash 2603StepFun | 17.00 | - | 0.10 | 0.30 | 218 tok/s |
| Gemini 3.5 Flash (high)Google | 33.00 | $2172 | 1.50 | 9.00 | 217 tok/s |
| Granite 4.2 3BIBMBest value frontier | 9.10 | $14.9 | 0.03 | 0.12 | 216 tok/s |
| gpt-oss-120b (high)OpenAI | 12.30 | $122 | 0.15 | 0.59 | 215 tok/s |
| Gemini 3 Flash Preview (Non-reasoning)Google | 17.90 | - | 0.50 | 3.00 | 214 tok/s |
| Gemini 3 Flash Preview (Reasoning)Google | 26.30 | - | 0.50 | 3.00 | 214 tok/s |
| Gemini 2.5 Flash (Non-reasoning)Google | 9.90 | - | 0.30 | 2.50 | 213 tok/s |
| o3-mini (high)OpenAI | 11.00 | - | 1.10 | 4.40 | 210 tok/s |
| NVIDIA Nemotron 3 Nano 30B A3B (Reasoning)NVIDIA | 8.90 | $63.7 | 0.05 | 0.20 | 205 tok/s |
| Gemini 3.6 Flash (high)Google | 34.30 | $1037 | 0.75 | 3.75 | 204 tok/s |
| o3-miniOpenAI | 12.50 | - | 1.10 | 4.40 | 204 tok/s |
| LFM2.5-2.6BLiquid AI | 8.40 | - | 0.00 | 0.00 | 200 tok/s |
| Qwen3.7 MaxAlibaba | 29.90 | $2277 | 2.50 | 7.50 | 199 tok/s |
| Muse Spark 1.3 (xhigh)Meta | 45.20 | $1655 | 1.25 | 4.25 | 198 tok/s |
| gpt-oss-20b (high)OpenAI | 9.00 | $29.4 | 0.06 | 0.19 | 198 tok/s |
| GPT-5 nano (minimal)OpenAI | 7.10 | - | 0.05 | 0.40 | 197 tok/s |
| Gemini 3.5 Flash (minimal)Google | 23.80 | - | 1.50 | 9.00 | 197 tok/s |
| Qwen3 Next 80B A3B (Reasoning)Alibaba | 11.20 | - | 0.15 | 1.20 | 193 tok/s |
| GPT-5 nano (high)OpenAI | 13.00 | - | 0.05 | 0.40 | 186 tok/s |
| Ministral 3 3BMistral | 4.80 | $61.3 | 0.10 | 0.10 | 186 tok/s |
| GPT-5.4 mini (medium)OpenAI | 19.70 | - | 0.75 | 4.50 | 184 tok/s |
| Nova 2.0 Lite (medium)Amazon | 12.50 | - | 0.30 | 2.50 | 181 tok/s |
| Qwen3 Next 80B A3B InstructAlibaba | 9.60 | - | 0.15 | 1.20 | 178 tok/s |
| Agnes 2.5 Pro AlphaSapiens AI | 26.80 | - | 0.45 | 0.90 | 178 tok/s |
| GPT-5 nano (medium)OpenAI | 12.50 | - | 0.05 | 0.40 | 177 tok/s |
| GPT-5.4 nano (xhigh)OpenAI | 21.20 | $331 | 0.20 | 1.25 | 176 tok/s |
| Nova LiteAmazon | 6.70 | - | 0.06 | 0.24 | 176 tok/s |
| o3OpenAI | 20.20 | - | 2.00 | 8.00 | 175 tok/s |
| GPT-5.4 nano (Non-Reasoning)OpenAI | 11.70 | - | 0.20 | 1.25 | 171 tok/s |
| GPT-5.4 mini (Non-Reasoning)OpenAI | 11.10 | - | 0.75 | 4.50 | 170 tok/s |
| GPT-5.4 nano (medium)OpenAI | 20.00 | - | 0.20 | 1.25 | 170 tok/s |
| Mistral Small 4 (Reasoning)Mistral | 11.50 | $78.7 | 0.15 | 0.60 | 168 tok/s |
| Nemotron 3 Ultra 550B A55B (Reasoning)NVIDIA | 23.40 | $801 | 0.60 | 2.60 | 168 tok/s |
| Quasar 438B (max, based on GLM-5.2)Multiverse Computing | 27.10 | $2504 | 0.60 | 1.80 | 164 tok/s |
| Nova 2.0 Lite (low)Amazon | 11.80 | - | 0.30 | 2.50 | 164 tok/s |
| Qwen3.5 35B A3B (Non-reasoning)Alibaba | 15.10 | - | 0.25 | 2.00 | 164 tok/s |
| Nova 2.0 Lite (high)Amazon | 13.40 | - | 0.30 | 2.50 | 161 tok/s |
| GPT-4.1OpenAI | 12.70 | - | 2.00 | 8.00 | 161 tok/s |
| Mistral Small (Sep '24)Mistral | 5.80 | - | 0.20 | 0.60 | 160 tok/s |
| GPT-4o (Nov '24)OpenAI | 8.40 | - | 2.50 | 10.00 | 159 tok/s |
| Mistral Small 3Mistral | 6.70 | - | 0.10 | 0.30 | 158 tok/s |
| Mistral Small 4 (Non-reasoning)Mistral | 9.00 | - | 0.15 | 0.60 | 158 tok/s |
| o4-mini (high)OpenAI | 16.70 | - | 1.10 | 4.40 | 157 tok/s |
| Mistral Small 3.2Mistral | 7.00 | $163 | 0.10 | 0.30 | 157 tok/s |
| Nova 2.0 Lite (Non-reasoning)Amazon | 8.70 | - | 0.30 | 2.50 | 157 tok/s |
| MiniMax-M3MiniMax | 29.60 | $538 | 0.30 | 1.20 | 156 tok/s |
| Mistral Small (Feb '24)Mistral | 5.50 | - | 0.15 | 0.60 | 152 tok/s |
| Mistral Small 3.1Mistral | 7.40 | $35.8 | 0.10 | 0.30 | 150 tok/s |
| Qwen3 30B A3B 2507 InstructAlibaba | 7.50 | - | 0.20 | 0.80 | 148 tok/s |
| Mistral Medium 3.5Mistral | 14.90 | $1160 | 1.50 | 7.50 | 148 tok/s |
| GPT-4o miniOpenAI | 6.70 | - | 0.15 | 0.60 | 148 tok/s |
| GPT-5.5 Instant (June 2026)OpenAI | 26.80 | $1184 | 5.00 | 30.00 | 147 tok/s |
| Qwen3.5 35B A3B (Reasoning)Alibaba | 19.30 | - | 0.25 | 2.00 | 146 tok/s |
| NVIDIA Nemotron Nano 9B V2 (Non-reasoning)NVIDIA | 6.80 | - | 0.05 | 0.20 | 146 tok/s |
| GPT-5.4 (xhigh)OpenAI | 39.00 | - | 2.50 | 15.00 | 144 tok/s |
| Qwen3.5 122B A10B (Non-reasoning)Alibaba | 17.70 | - | 0.40 | 3.20 | 144 tok/s |
| Qwen3.6 35B A3B (Non-reasoning)Alibaba | 15.20 | - | 0.38 | 2.25 | 144 tok/s |
| Solar Pro 3Upstage | 7.80 | $159 | 0.15 | 0.60 | 143 tok/s |
| Qwen3 30B A3B 2507 (Reasoning)Alibaba | 9.80 | $188 | 0.20 | 2.40 | 141 tok/s |
| Mistral Medium 3.1Mistral | 9.90 | - | 0.40 | 2.00 | 141 tok/s |
| Llama 3.1 Instruct 8BMeta | 6.90 | - | 0.02 | 0.05 | 141 tok/s |
| Mistral Medium 3Mistral | 9.00 | - | 0.40 | 2.00 | 140 tok/s |
| Mistral MediumMistral | 5.50 | - | 1.50 | 7.50 | 137 tok/s |
| GPT-5.3 Codex (xhigh)OpenAI | 32.50 | - | 1.75 | 14.00 | 137 tok/s |
| GPT-4.1 nanoOpenAI | 7.80 | - | 0.10 | 0.40 | 133 tok/s |
| Grok 4.3 (Non-reasoning)SpaceXAI | 14.50 | $203 | 1.25 | 2.50 | 132 tok/s |
| Qwen3.5 122B A10B (Reasoning)Alibaba | 16.20 | $661 | 0.40 | 3.20 | 129 tok/s |
| DeepSeek V4 Flash 0731 (Reasoning, Max Effort)DeepSeek | 34.50 | $474 | 0.44 | 1.32 | 128 tok/s |
| Gemini 2.5 ProGoogle | 16.70 | $707 | 1.25 | 10.00 | 127 tok/s |
| Inkling SmallThinking Machines | 26.10 | $241 | 0.30 | 1.20 | 126 tok/s |
| Grok 4.3 (high)SpaceXAI | 25.40 | $332 | 1.25 | 2.50 | 123 tok/s |
| Nova 2.0 Pro Preview (medium)Amazon | 14.20 | - | 1.25 | 10.00 | 123 tok/s |
| Ring-2.6-1TInclusionAI | 17.30 | $576 | 0.30 | 2.50 | 121 tok/s |
| GPT-4o (Aug '24)OpenAI | 7.70 | - | 2.50 | 10.00 | 121 tok/s |
| GPT-5.1 (high)OpenAI | 24.70 | - | 1.25 | 10.00 | 121 tok/s |
| Kimi K2 ThinkingKimi | 22.00 | - | 0.60 | 2.50 | 121 tok/s |
| DeepSeek V4 Flash Vision (Reasoning, Max Effort)DeepSeek | 35.00 | $445 | 0.44 | 1.32 | 121 tok/s |
| Grok 4.3 (low)SpaceXAI | 24.30 | - | 1.25 | 2.50 | 121 tok/s |
| Nova 2.0 Pro Preview (low)Amazon | 12.80 | - | 1.25 | 10.00 | 120 tok/s |
| Qwen3.6 35B A3B (Reasoning)Alibaba | 22.30 | - | 0.38 | 2.25 | 120 tok/s |
| Granite 4.2 8BIBM | 12.40 | - | 0.06 | 0.25 | 120 tok/s |
| GPT-5.6 Luna (max)OpenAI | 37.50 | $320 | 0.20 | 1.20 | 120 tok/s |
| Qwen3 VL 8B InstructAlibaba | 7.30 | - | 0.18 | 0.70 | 119 tok/s |
| Gemini 3.1 Pro PreviewGoogle | 30.40 | $1310 | 2.00 | 12.00 | 118 tok/s |
| GPT-5.6 Luna (Non-reasoning)OpenAI | 16.80 | - | 0.20 | 1.20 | 118 tok/s |
| GPT-4o (May '24)OpenAI | 7.30 | - | 5.00 | 15.00 | 117 tok/s |
| GLM-5.2 (Non-reasoning)Z AI | 22.40 | - | 1.40 | 4.40 | 117 tok/s |
| Qwen3 Coder NextAlibaba | 10.10 | $476 | 0.35 | 1.20 | 117 tok/s |
| GPT-5.6 Terra (max)OpenAI | 42.30 | $2501 | 2.00 | 12.00 | 116 tok/s |
| GPT-5.6 Luna (high)OpenAI | 32.90 | - | 0.20 | 1.20 | 116 tok/s |
| GPT-5.6 Luna (low)OpenAI | 21.80 | - | 0.20 | 1.20 | 116 tok/s |
| Nex-N2-ProNex AGI | 28.20 | - | 0.50 | 2.50 | 115 tok/s |
| Grok 4.3 (medium)SpaceXAI | 24.80 | - | 1.25 | 2.50 | 114 tok/s |
| GPT-4.1 miniOpenAI | 10.20 | - | 0.40 | 1.60 | 114 tok/s |
| Qwen3 VL 8B (Reasoning)Alibaba | 8.20 | - | 0.18 | 2.10 | 113 tok/s |
| Qwen3 VL 30B A3B InstructAlibaba | 7.90 | - | 0.20 | 0.80 | 110 tok/s |
| GPT-5.6 Luna (xhigh)OpenAIBest value frontier | 34.80 | $179 | 0.20 | 1.20 | 110 tok/s |
| Qwen3 VL 30B A3B (Reasoning)Alibaba | 9.50 | - | 0.20 | 2.40 | 110 tok/s |
| NVIDIA Nemotron Nano 9B V2 (Reasoning)NVIDIA | 7.40 | - | 0.04 | 0.16 | 110 tok/s |
| GPT-5.6 Luna (medium)OpenAI | 25.80 | - | 0.20 | 1.20 | 110 tok/s |
| Qwen2.5 TurboAlibaba | 6.40 | - | 0.05 | 0.20 | 109 tok/s |
| GPT-5.4 (low)OpenAI | 27.60 | - | 2.50 | 15.00 | 109 tok/s |
| Nova 2.0 Pro Preview (Non-reasoning)Amazon | 10.00 | - | 1.25 | 10.00 | 108 tok/s |
| GPT-5.4 (Non-reasoning)OpenAI | 18.20 | - | 2.50 | 15.00 | 108 tok/s |
| Gemma 4 12B (Reasoning)Google | 14.20 | - | 0.10 | 0.30 | 108 tok/s |
| Gemma 4 12B (Non-reasoning)Google | 9.40 | - | 0.10 | 0.30 | 107 tok/s |
| Qwen3 30B A3B (Non-reasoning)Alibaba | 6.60 | - | 0.20 | 0.80 | 107 tok/s |
| GPT-5 mini (high)OpenAI | 17.40 | $189 | 0.25 | 2.00 | 107 tok/s |
| GLM-4.7 (Non-reasoning)Z AI | 17.40 | - | 0.60 | 2.20 | 106 tok/s |
| GPT-5.1 (Non-reasoning)OpenAI | 13.30 | - | 1.25 | 10.00 | 106 tok/s |
| GLM-4.7-Flash (Reasoning)Z AI | 14.90 | - | 0.07 | 0.40 | 105 tok/s |
| Qwen3 32B (Non-reasoning)Alibaba | 7.30 | - | 0.16 | 0.64 | 105 tok/s |
| Qwen3 30B A3B (Reasoning)Alibaba | 7.60 | - | 0.20 | 2.40 | 104 tok/s |
| GPT-5 mini (minimal)OpenAI | 9.90 | - | 0.25 | 2.00 | 104 tok/s |
| Muse Glimmer (high)Meta | 18.10 | $152 | 0.35 | 1.50 | 104 tok/s |
| Qwen3 32B (Reasoning)Alibaba | 7.20 | - | 0.16 | 0.64 | 104 tok/s |
| Grok 4.20 0309 v2 (Reasoning)SpaceXAI | 25.70 | - | 1.25 | 2.50 | 103 tok/s |
| GPT-5 mini (medium)OpenAI | 20.60 | - | 0.25 | 2.00 | 102 tok/s |
| GPT-5 (high)OpenAI | 23.00 | - | 1.25 | 10.00 | 101 tok/s |
| Qwen3 Omni 30B A3B (Reasoning)Alibaba | 7.80 | - | 0.25 | 0.97 | 100 tok/s |
| Qwen3.5 Omni PlusAlibaba | 20.40 | - | 0.40 | 4.80 | 100 tok/s |
| GPT-5.6 Terra (xhigh)OpenAI | 38.20 | $1187 | 2.00 | 12.00 | 99 tok/s |
| Nemotron 3 Super 120B A12B (Reasoning)NVIDIA | 13.60 | $859 | 0.19 | 0.65 | 99 tok/s |
| GLM-4.7 (Reasoning)Z AI | 22.20 | - | 0.60 | 2.20 | 98 tok/s |
| GPT-5 (medium)OpenAI | 22.90 | - | 1.25 | 10.00 | 97 tok/s |
| MiniMax-M2.5MiniMax | 22.80 | - | 0.30 | 1.20 | 96 tok/s |
| Qwen3 Omni 30B A3B InstructAlibaba | 6.00 | - | 0.25 | 0.97 | 95 tok/s |
| GPT-5.6 Terra (high)OpenAI | 34.50 | $768 | 2.00 | 12.00 | 95 tok/s |
| Hy3Tencent | 25.80 | $149 | 0.14 | 0.55 | 94 tok/s |
| MiniMax-M2MiniMax | 18.60 | - | 0.30 | 1.20 | 94 tok/s |
| GLM-4.7-Flash (Non-reasoning)Z AI | 10.60 | - | 0.07 | 0.40 | 94 tok/s |
| GPT-5.6 Terra (medium)OpenAI | 32.80 | - | 2.00 | 12.00 | 94 tok/s |
| GPT-5.6 Terra (Non-reasoning)OpenAI | 22.30 | - | 2.00 | 12.00 | 94 tok/s |
| GPT-5.6 Terra (low)OpenAI | 27.90 | - | 2.00 | 12.00 | 93 tok/s |
| Llama 4 ScoutMeta | 6.50 | $77.6 | 0.19 | 0.68 | 93 tok/s |
| Llama 4 MaverickMeta | 9.30 | $45.5 | 0.26 | 0.91 | 93 tok/s |
| Grok 4.20 0309 v2 (Non-reasoning)SpaceXAI | 14.20 | - | 1.25 | 2.50 | 92 tok/s |
| Ministral 3 14BMistral | 6.00 | $250 | 0.20 | 0.20 | 92 tok/s |
| MiniMax-M2.1MiniMax | 20.90 | - | 0.30 | 1.20 | 91 tok/s |
| GPT-5.5 (medium)OpenAI | 34.20 | $2095 | 5.00 | 30.00 | 91 tok/s |
| GPT-5.5 (xhigh)OpenAI | 38.60 | $5294 | 5.00 | 30.00 | 90 tok/s |
| Qwen3 Coder 30B A3B InstructAlibaba | 9.60 | - | 0.45 | 2.25 | 90 tok/s |
| Qwen3 VL 32B (Reasoning)Alibaba | 11.90 | - | 0.16 | 0.64 | 90 tok/s |
| Ministral 3 8BMistral | 5.50 | $169 | 0.15 | 0.15 | 89 tok/s |
| GPT-5.2 (xhigh)OpenAI | 30.40 | - | 1.75 | 14.00 | 89 tok/s |
| Llama 3.3 Instruct 70BMeta | 7.70 | - | 0.66 | 0.72 | 88 tok/s |
| Qwen3.5 397B A17B (Reasoning)Alibaba | 19.10 | $941 | 0.60 | 3.60 | 86 tok/s |
| Qwen3.5 9B (Non-reasoning)Alibaba | 13.30 | - | 0.17 | 0.25 | 85 tok/s |
| GPT-5.5 (Non-reasoning)OpenAI | 23.20 | - | 5.00 | 30.00 | 84 tok/s |
| Qwen3.5 27B (Non-reasoning)Alibaba | 19.40 | - | 0.30 | 2.40 | 84 tok/s |
| Claude 4.5 Haiku (Reasoning)Anthropic | 17.60 | $524 | 1.00 | 5.00 | 83 tok/s |
| Qwen3.5 397B A17B (Non-reasoning)Alibaba | 21.40 | - | 0.60 | 3.60 | 83 tok/s |
| Qwen3.5 9B (Reasoning)Alibaba | 13.70 | - | 0.14 | 0.20 | 83 tok/s |
| Gemma 4 26B A4B (Non-reasoning)Google | 13.10 | - | 0.13 | 0.40 | 82 tok/s |
| GPT-5 (minimal)OpenAI | 11.40 | - | 1.25 | 10.00 | 81 tok/s |
| GPT-5.5 (high)OpenAI | 37.30 | $3411 | 5.00 | 30.00 | 81 tok/s |
| Step 3.7 FlashStepFun | 19.50 | - | 0.20 | 1.15 | 81 tok/s |
| Claude 4.5 Haiku (Non-reasoning)Anthropic | 15.40 | - | 1.00 | 5.00 | 81 tok/s |
| Solar Pro 4Upstage | 28.20 | - | 0.30 | 1.20 | 80 tok/s |
| GPT-5.5 (low)OpenAI | 30.70 | - | 5.00 | 30.00 | 80 tok/s |
| DeepSeek V4 Pro 0813 (Reasoning, Max Effort)DeepSeek | 36.30 | $1122 | 1.32 | 3.96 | 80 tok/s |
| Claude Sonnet 5 (Adaptive Reasoning, Max Effort)Anthropic | 38.40 | $6998 | 2.00 | 10.00 | 79 tok/s |
| Mistral Large 3Mistral | 9.70 | $130 | 0.50 | 1.50 | 79 tok/s |
| Grok 3 mini Reasoning (high)SpaceXAI | 14.60 | - | 0.30 | 0.50 | 79 tok/s |
| GPT-5 (low)OpenAI | 20.80 | - | 1.25 | 10.00 | 78 tok/s |
| Qwen3.5 27B (Reasoning)Alibaba | 22.90 | - | 0.30 | 2.40 | 77 tok/s |
| DeepSeek V4 Pro (Reasoning, Max Effort)DeepSeek | 30.90 | $277 | 0.43 | 0.87 | 77 tok/s |
| Mistral 7B InstructMistral | 5.00 | - | 0.25 | 0.25 | 76 tok/s |
| Granite 4.2 30BIBM | 14.80 | - | 0.16 | 0.65 | 76 tok/s |
| GPT-5.2 (Non-reasoning)OpenAI | 17.00 | - | 1.75 | 14.00 | 75 tok/s |
| Grok Build 0.1 0616SpaceXAI | 27.20 | - | 1.00 | 2.00 | 73 tok/s |
| MiniMax-M2.7MiniMax | 23.20 | $236 | 0.30 | 1.20 | 73 tok/s |
| DeepSeek V4 Pro (Non-reasoning)DeepSeek | 20.80 | - | 0.43 | 0.87 | 73 tok/s |
| Qwen3.8-Flash-NextAlibaba | 42.20 | - | 0.15 | 0.47 | 72 tok/s |
| Inkling (xhigh)Thinking Machines | 25.50 | $1026 | 1.00 | 4.05 | 72 tok/s |
| GLM-5 (Reasoning)Z AI | 27.90 | - | 1.00 | 3.20 | 71 tok/s |
| Kimi K2.7 CodeKimi | 26.30 | $993 | 0.95 | 4.00 | 71 tok/s |
| GLM-5.3 (max)Z AI | 44.90 | $2503 | 1.40 | 4.40 | 70 tok/s |
| Granite 4.1 8BIBM | 6.60 | - | 0.05 | 0.10 | 70 tok/s |
| DeepSeek V4 Pro (Reasoning, High Effort)DeepSeek | 30.10 | - | 0.43 | 0.87 | 70 tok/s |
| NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning)NVIDIA | 5.80 | - | 0.20 | 0.60 | 70 tok/s |
| Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback)AnthropicBest value frontier | 53.40 | $13129 | 10.00 | 50.00 | 69 tok/s |
| GLM-4.6 (Non-reasoning)Z AI | 14.90 | - | 0.57 | 2.20 | 69 tok/s |
| Claude Sonnet 5 (Adaptive Reasoning, High Effort)Anthropic | - | - | 2.00 | 10.00 | 69 tok/s |
| Claude Sonnet 5 (Adaptive Reasoning, Xhigh Effort)Anthropic | - | - | 2.00 | 10.00 | 68 tok/s |
| Qwen3 VL 32B InstructAlibaba | 8.40 | - | 0.16 | 0.64 | 68 tok/s |
| GPT-5.6 Sol (max)OpenAI | 47.10 | $3465 | 4.00 | 20.00 | 68 tok/s |
| Command ACohere | 7.00 | - | 2.50 | 10.00 | 67 tok/s |
| GPT-5.6 Sol (low)OpenAI | 33.80 | $637 | 4.00 | 20.00 | 67 tok/s |
| GLM-5.3-FlashZ AIBest value frontier | 41.90 | $280 | 0.15 | 0.50 | 66 tok/s |
| MiMo-V2.5XiaomiBest value frontier | 22.30 | $42.6 | 0.14 | 0.28 | 65 tok/s |
| GPT-5.6 Sol (Non-reasoning)OpenAI | 28.30 | - | 4.00 | 20.00 | 65 tok/s |
| GPT-5.6 Sol (medium)OpenAI | 39.50 | $997 | 4.00 | 20.00 | 65 tok/s |
| GLM-4.5-AirZ AI | 11.10 | - | 0.17 | 0.98 | 64 tok/s |
| Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)Anthropic | 49.70 | $11161 | 10.00 | 50.00 | 64 tok/s |
| Claude Sonnet 5 (Adaptive Reasoning, Medium Effort)Anthropic | 28.40 | $1218 | 2.00 | 10.00 | 64 tok/s |
| GLM-5.2 (max)Z AI | 38.60 | - | 1.40 | 4.40 | 64 tok/s |
| GPT-5.6 Sol (xhigh)OpenAI | 44.10 | $2082 | 4.00 | 20.00 | 63 tok/s |
| GLM-4.6V (Reasoning)Z AI | 11.20 | - | 0.30 | 0.90 | 63 tok/s |
| Claude Sonnet 5 (Adaptive Reasoning, Low Effort)Anthropic | 24.70 | $653 | 2.00 | 10.00 | 62 tok/s |
| GPT-5.6 Sol (high)OpenAIBest value frontier | 42.50 | $1487 | 4.00 | 20.00 | 62 tok/s |
| Claude Sonnet 5 (Non-reasoning, High Effort)Anthropic | 28.90 | - | 2.00 | 10.00 | 61 tok/s |
| Qwen3 235B A22B (Non-reasoning)Alibaba | 8.30 | - | 0.70 | 2.80 | 61 tok/s |
| Grok 4.6 (medium)SpaceXAI | 43.00 | $1937 | 2.00 | 6.00 | 61 tok/s |
| Qwen3 14B (Non-reasoning)Alibaba | 6.70 | - | 0.35 | 1.40 | 60 tok/s |
| Gemma 4 31B (Non-reasoning)Google | 13.90 | - | 0.14 | 0.40 | 60 tok/s |
| Grok 4.6 (low)SpaceXAI | 35.40 | $761 | 2.00 | 6.00 | 60 tok/s |
| Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback)AnthropicBest value frontier | 53.20 | $9063 | 10.00 | 50.00 | 60 tok/s |
| Grok 4.6 (xhigh)SpaceXAI | 44.30 | $2830 | 2.00 | 6.00 | 60 tok/s |
| Llama 3.1 Instruct 70BMeta | 6.60 | - | 0.56 | 0.56 | 60 tok/s |
| Qwen3 Max Thinking (Preview)Alibaba | 16.30 | - | 1.20 | 6.00 | 59 tok/s |
| Claude Opus 4.8 (Adaptive Reasoning, Max Effort)Anthropic | 42.00 | $6874 | 5.00 | 25.00 | 59 tok/s |
| Grok 4.5 (high)SpaceXAI | 39.10 | $1404 | 2.00 | 6.00 | 59 tok/s |
| GLM-5 (Non-reasoning)Z AI | 21.80 | - | 1.00 | 3.20 | 59 tok/s |
| Qwen3 235B A22B 2507 InstructAlibaba | 12.00 | - | 0.23 | 0.92 | 59 tok/s |
| Grok 4.6 (high)SpaceXAI | 44.40 | $2352 | 2.00 | 6.00 | 58 tok/s |
| Qwen3 235B A22B (Reasoning)Alibaba | 9.50 | - | 0.70 | 8.40 | 58 tok/s |
| Qwen3 14B (Reasoning)Alibaba | 6.40 | - | 0.35 | 4.20 | 58 tok/s |
| Qwen3 Max (Preview)Alibaba | 12.60 | - | 1.20 | 6.00 | 58 tok/s |
| Qwen3.6 27B (Non-reasoning)Alibaba | 19.80 | - | 0.60 | 3.60 | 58 tok/s |
| Qwen3 235B A22B 2507 (Reasoning)Alibaba | 12.70 | $235 | 0.23 | 2.30 | 57 tok/s |
| Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback)Anthropic | 51.20 | $5242 | 10.00 | 50.00 | 57 tok/s |
| GLM-5.1 (Reasoning)Z AI | 27.40 | - | 1.20 | 4.40 | 57 tok/s |
| Qwen3 Coder 480B A35B InstructAlibaba | 11.90 | - | 1.50 | 7.50 | 57 tok/s |
| Qwen3.6 27B (Reasoning)Alibaba | 21.90 | $1145 | 0.60 | 3.60 | 57 tok/s |
| GPT-6 Astra (max)OpenAIBest value frontier | 52.80 | $5324 | 10.00 | 50.00 | 56 tok/s |
| Kimi K2.6Kimi | 31.30 | - | 0.95 | 4.00 | 56 tok/s |
| Qwen3.6 Max PreviewAlibaba | 28.40 | - | 1.30 | 7.80 | 56 tok/s |
| Qwen3.6 PlusAlibaba | 27.00 | - | 0.50 | 3.00 | 56 tok/s |
| Qwen3.7 PlusAlibaba | 25.80 | $568 | 0.40 | 1.60 | 56 tok/s |
| Qwen3.8 27B (Non-reasoning)Alibaba | 22.40 | - | 0.50 | 3.00 | 56 tok/s |
| Qwen3 VL 235B A22B (Reasoning)Alibaba | 13.40 | - | 0.40 | 4.00 | 56 tok/s |
| Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback)Anthropic | 49.10 | $3983 | 10.00 | 50.00 | 55 tok/s |
| Qwen3.8 27B (low)Alibaba | 29.20 | - | 0.50 | 3.00 | 54 tok/s |
| Claude Opus 5 (Adaptive Reasoning, Max Effort)Anthropic | 50.70 | $7275 | 5.00 | 25.00 | 54 tok/s |
| Qwen3.8 27B (medium)Alibaba | 30.70 | - | 0.50 | 3.00 | 54 tok/s |
| Claude Opus 5 (Adaptive Reasoning, Medium Effort)Anthropic | 45.10 | $2732 | 5.00 | 25.00 | 53 tok/s |
| Qwen3 MaxAlibaba | 15.60 | - | 1.20 | 6.00 | 53 tok/s |
| Claude Opus 5 (Adaptive Reasoning, Low Effort)Anthropic | 39.80 | $1561 | 5.00 | 25.00 | 53 tok/s |
| GLM-4.6 (Reasoning)Z AI | 18.50 | - | 0.55 | 2.20 | 53 tok/s |
| GPT-6 Astra (xhigh)OpenAIBest value frontier | 52.50 | $3803 | 10.00 | 50.00 | 52 tok/s |
| Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)Anthropic | 49.70 | $5868 | 5.00 | 25.00 | 52 tok/s |
| Claude Opus 4.5 (Reasoning)Anthropic | 29.10 | - | 5.00 | 25.00 | 52 tok/s |
| Claude Fable 5.1 (Adaptive Reasoning, Low Effort, Default Fallback)Anthropic | 47.00 | $3158 | 10.00 | 50.00 | 52 tok/s |
| Claude Opus 5 (Adaptive Reasoning, High Effort)Anthropic | 48.20 | $4332 | 5.00 | 25.00 | 52 tok/s |
| GLM-4.5V (Non-reasoning)Z AI | 6.70 | - | 0.60 | 1.80 | 51 tok/s |
| Kimi K2.6 (Non-reasoning)Kimi | 23.60 | - | 0.95 | 4.00 | 51 tok/s |
| Qwen3 VL 235B A22B InstructAlibaba | 9.90 | - | 0.40 | 1.60 | 51 tok/s |
| Claude Opus 4.7 (Adaptive Reasoning, Max Effort)Anthropic | 40.70 | - | 5.00 | 25.00 | 50 tok/s |
| GPT-6 Astra (high)OpenAIBest value frontier | 51.00 | $2917 | 10.00 | 50.00 | 50 tok/s |
| Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort)Anthropic | 30.50 | $4925 | 3.00 | 15.00 | 50 tok/s |
| LongCat 2.0LongCat | 19.70 | $545 | 0.75 | 2.95 | 50 tok/s |
| Claude Opus 4.5 (Non-reasoning)Anthropic | 23.70 | - | 5.00 | 25.00 | 48 tok/s |
| GPT-6 Astra (medium)OpenAIBest value frontier | 49.70 | $2434 | 10.00 | 50.00 | 47 tok/s |
| GPT-6 Astra (low)OpenAIBest value frontier | 46.00 | $1537 | 10.00 | 50.00 | 47 tok/s |
| Qwen3.8 27B (xhigh)Alibaba | 33.90 | $1170 | 0.50 | 3.00 | 47 tok/s |
| Claude Opus 4.7 (Non-reasoning, High Effort)Anthropic | 30.90 | - | 5.00 | 25.00 | 46 tok/s |
| GLM-4.5V (Reasoning)Z AI | 7.60 | - | 0.60 | 1.80 | 46 tok/s |
| Claude 4.5 Sonnet (Reasoning)Anthropic | 21.20 | $1501 | 3.00 | 15.00 | 46 tok/s |
| Phi-4 Mini InstructMicrosoft | 6.30 | - | 0.00 | 0.00 | 45 tok/s |
| North Mini CodeCohere | 12.80 | - | 0.00 | 0.00 | 45 tok/s |
| Claude Sonnet 4.6 (Non-reasoning, Low Effort)Anthropic | 23.30 | - | 3.00 | 15.00 | 45 tok/s |
| GLM-4.6V (Non-reasoning)Z AI | 8.40 | - | 0.30 | 0.90 | 44 tok/s |
| Claude Sonnet 4.6 (Non-reasoning, High Effort)Anthropic | 24.70 | - | 3.00 | 15.00 | 44 tok/s |
| Hermes 4 - Llama-3.1 405B (Non-reasoning)Nous Research | 7.40 | - | 1.00 | 3.00 | 43 tok/s |
| Hermes 4 - Llama-3.1 405B (Reasoning)Nous Research | 7.50 | - | 1.00 | 3.00 | 42 tok/s |
| Claude 4.5 Sonnet (Non-reasoning)Anthropic | 19.30 | - | 3.00 | 15.00 | 42 tok/s |
| Kimi K3 (max)Kimi | 43.80 | $3658 | 3.00 | 15.00 | 42 tok/s |
| Gemma 4 E4B (Reasoning)Google | 8.90 | - | 0.02 | 0.10 | 41 tok/s |
| Llama Nemotron Super 49B v1.5 (Reasoning)NVIDIA | 9.00 | - | 0.40 | 0.40 | 41 tok/s |
| Phi-4Microsoft | 5.90 | - | 0.13 | 0.50 | 41 tok/s |
| GLM-5.1 (Non-reasoning)Z AI | 24.20 | - | 1.38 | 4.40 | 41 tok/s |
| Kimi K2Kimi | 12.70 | - | 0.57 | 2.30 | 40 tok/s |
| Qwen3.8 2.4T A95BAlibaba | 40.00 | $2632 | 2.00 | 6.00 | 40 tok/s |
| Kimi K2 0905Kimi | 15.30 | - | 0.60 | 2.50 | 40 tok/s |
| Gemma 4 E4B (Non-reasoning)Google | 7.50 | - | 0.02 | 0.10 | 40 tok/s |
| Qwen3.8 MaxAlibaba | 40.30 | $3172 | 2.00 | 6.00 | 39 tok/s |
| Claude Opus 4.6 (Adaptive Reasoning, Max Effort)Anthropic | 31.90 | - | 5.00 | 25.00 | 39 tok/s |
| Qwen3 8B (Non-reasoning)Alibaba | 6.00 | - | 0.18 | 0.70 | 39 tok/s |
| Claude Opus 4.6 (Non-reasoning, High Effort)Anthropic | 26.40 | - | 5.00 | 25.00 | 39 tok/s |
| Qwen3 8B (Reasoning)Alibaba | 5.20 | - | 0.18 | 2.10 | 39 tok/s |
| Kimi K3 (low)Kimi | 34.50 | - | 3.00 | 15.00 | 38 tok/s |
| MiMo-V2.5-Pro (Non-reasoning)Xiaomi | 18.30 | - | 0.43 | 0.87 | 36 tok/s |
| MiMo-V2.5-ProXiaomiBest value frontier | 26.40 | $138 | 0.43 | 0.87 | 36 tok/s |
| Llama Nemotron Super 49B v1.5 (Non-reasoning)NVIDIA | 7.40 | - | 0.40 | 0.40 | 35 tok/s |
| Nova PremierAmazon | 9.20 | - | 2.50 | 12.50 | 33 tok/s |
| GPT-4 TurboOpenAI | 7.00 | - | 10.00 | 30.00 | 32 tok/s |
| NVIDIA Nemotron Nano 12B v2 VL (Reasoning)NVIDIA | 7.50 | - | 0.20 | 0.60 | 32 tok/s |
| Llama 3.1 Nemotron Instruct 70BNVIDIA | 6.90 | - | 1.20 | 1.20 | 31 tok/s |
| Seed-OSS-36B-InstructByteDance Seed | 12.10 | - | 0.21 | 0.57 | 30 tok/s |
| Hermes 3 - Llama-3.1 70BNous Research | 6.00 | - | 0.70 | 0.70 | 29 tok/s |
| Qwen3.5 4B (Reasoning)Alibaba | 13.10 | - | 0.03 | 0.15 | 27 tok/s |
| Qwen3.5 4B (Non-reasoning)Alibaba | 10.80 | - | 0.03 | 0.15 | 22 tok/s |
| Granite 4.0 H SmallIBM | 6.00 | - | 0.06 | 0.25 | 22 tok/s |
| DeepSeek R1 Distill Llama 70BDeepSeek | 7.90 | - | 0.70 | 1.10 | 21 tok/s |
| Phi-4 Multimodal InstructMicrosoft | 5.80 | - | 0.00 | 0.00 | 17 tok/s |
| Granite 3.3 8B (Non-reasoning)IBM | 4.90 | - | 0.03 | 0.25 | 16 tok/s |
| Llama 3.2 Instruct 11B (Vision)Meta | 5.40 | - | 0.34 | 0.34 | 12 tok/s |
| Ling-flash-2.0InclusionAI | 7.80 | - | 0.14 | 0.57 | 7 tok/s |
Guides
How we compare models
Intelligence uses the Artificial Analysis Intelligence Index. Benchmark cost is the total USD cost of running that test, including reasoning tokens; it is not a per-token API price. Rankings use the same dataset and calculations in every language. General benchmark scores do not measure performance in a particular language or task.