Artificial Analysis Benchmark
275 models · 126 linked to model directory · Updated Sep 23, 2026 at 08:35 AM
| Model | Provider | Intelligence | Context | Price ($/M tok) | Speed (tok/s) | Latency (s) | Matched |
|---|---|---|---|---|---|---|---|
| Claude Opus 5.5 (max with fallback) | Anthropic | 58 | 1M | - | - | - | anthropic |
| Claude Opus 5.5 (xhigh with fallback) | Anthropic | 56 | 1M | - | 76 | 165.2 | — |
| Claude Opus 5.5 (high with fallback) | Anthropic | 54 | 1M | - | 90 | 12.8 | anthropic |
| Claude Fable 5.1 (max with fallback) | Anthropic | 53 | 1M | - | 65 | 284.4 | anthropic |
| Claude Fable 5.1 (xhigh with fallback) | Anthropic | 53 | 1M | - | 61 | 125.8 | — |
| GPT-6 Astra (max) | OpenAI | 53 | 1M | - | 54 | 323.7 | openai |
| GPT-6 Astra (xhigh) | OpenAI | 52 | 1M | - | 54 | 203.1 | — |
| Claude Opus 5.5 (medium with fallback) | Anthropic | 51 | 1M | - | 76 | 22.8 | anthropic |
| Claude Fable 5.1 (high with fallback) | Anthropic | 51 | 1M | - | 56 | 26.2 | anthropic |
| GPT-6 Astra (high) | OpenAI | 51 | 1M | - | 49 | 79.0 | openai |
| GPT-6 Astra (medium) | OpenAI | 50 | 1M | - | 47 | 6.4 | openai |
| Claude Fable 5.1 (medium with fallback) | Anthropic | 49 | 1M | - | 56 | 10.3 | anthropic |
| Muse Spark 1.3 (max) | Meta | 48 | 1M | - | 221 | 21.0 | meta |
| GPT-6 Sol (max) | OpenAI | 48 | 872k | - | 115 | 102.2 | openai |
| Claude Fable 5.1 (low with fallback) | Anthropic | 47 | 1M | - | 56 | 6.7 | anthropic |
| Grok 4.7 (xhigh) | SpaceXAI | 46 | 500k | - | 39 | 0.9 | xai |
| Grok 4.7 (high) | SpaceXAI | 46 | 500k | - | 53 | 0.8 | xai |
| MiMo-V2.6-Pro | Xiaomi | 46 | 1M | - | 58 | 2.6 | xiaomi |
| GPT-6 Astra (low) | OpenAI | 46 | 1M | - | 50 | 2.9 | openai |
| Qwen3.8 Max (0902) | Alibaba | 45 | 984k | - | 39 | 2.9 | alibaba |
| Muse Spark 1.3 (xhigh) | Meta | 45 | 1M | - | 246 | 23.5 | — |
| GLM-5.3 (max) | Z AI | 45 | 1M | - | 57 | 3.3 | zai |
| Grok 4.6 (high) | SpaceXAI | 44 | 500k | - | 59 | 42.8 | xai |
| Grok 4.6 (xhigh) | SpaceXAI | 44 | 500k | - | 56 | 51.9 | — |
| GPT-6 Sol (xhigh) | OpenAI | 44 | 872k | - | 128 | 46.5 | — |
| Step 5 Preview | StepFun | 44 | 1M | - | 84 | 4.1 | — |
| Kimi K3 (max) | Kimi | 44 | 1.05M | - | 37 | 4.2 | moonshotai |
| Grok 4.6 (medium) | SpaceXAI | 43 | 500k | - | 59 | 29.3 | xai |
| GPT-6 Sol (high) | OpenAI | 43 | 872k | - | 119 | 9.9 | openai |
| Claude Opus 5.5 (low with fallback) | Anthropic | 42 | 1M | - | 83 | 9.3 | anthropic |
| GPT-5.6 Terra (max) | OpenAI | 42 | 1M | - | 85 | 213.8 | openai |
| GLM-5.3-Flash | Z AI | 42 | 1M | - | 64 | 3.1 | zai |
| Gemini 3.8 Flash (high) | 41 | 1M | - | 297 | 14.3 | ||
| Qwen3.8 2.4T A95B | Alibaba | 40 | 984k | - | 38 | 3.0 | hyper |
| Qwen3.8-Flash-Next | Alibaba | 40 | 256k | - | 54 | 2.7 | vultr |
| GPT-6 Sol (medium) | OpenAI | 40 | 872k | - | 114 | 2.0 | openai |
| Gemini 3.8 Flash (medium) | 40 | 1M | - | - | - | ||
| DeepSeek V4.1 Flash (max) | DeepSeek | 39 | 1M | - | 221 | 1.0 | empiriolabs |
| Claude Sonnet 5 (max) | Anthropic | 38 | 1M | - | 81 | 185.3 | anthropic |
| GPT-5.6 Terra (xhigh) | OpenAI | 38 | 1M | - | 85 | 34.5 | — |
| GPT-6 Luna (max) | OpenAI | 37 | 1M | - | 154 | 124.2 | openai |
| DeepSeek V4 Pro 0813 (max) | DeepSeek | 36 | 1M | - | 68 | 1.7 | deepseek |
| Agnes 3.0 Flash | Sapiens AI | 36 | 1M | - | - | - | nano-gpt |
| Agnes 2.5 Pro Beta | Sapiens AI | 35 | 1M | - | - | - | — |
| Grok 4.6 (low) | SpaceXAI | 35 | 500k | - | 56 | 8.0 | xai |
| DeepSeek V4 Flash Vision (max) | DeepSeek | 35 | 1M | - | 231 | 1.1 | — |
| Kimi K3 (low) | Kimi | 34 | 1.05M | - | 37 | 5.1 | moonshotai |
| Claude Sonnet 5 (xhigh) | Anthropic | 34 | 1M | - | 66 | 35.3 | — |
| GPT-5.6 Terra (high) | OpenAI | 34 | 1M | - | 79 | 3.6 | openai |
| GPT-6 Sol (low) | OpenAI | 34 | 872k | - | 129 | 1.3 | openai |
| GPT-6 Luna (xhigh) | OpenAI | 34 | 1M | - | 153 | 22.5 | — |
| Qwen3.8 27B (xhigh) | Alibaba | 34 | 256k | - | 41 | 4.0 | empiriolabs |
| Motif 3 | Motif Technologies | 34 | 262k | - | - | - | — |
| Gemini 3.8 Flash (low) | 33 | 1M | - | - | - | ||
| GPT-5.3 Codex (xhigh) | OpenAI | 33 | 400k | - | 146 | 53.6 | openai |
| Motif 3 (Beta) | Motif Technologies | 32 | 262k | - | - | - | — |
| GPT-6 Luna (high) | OpenAI | 32 | 1M | - | 144 | 7.8 | openai |
| Claude Sonnet 5 (high) | Anthropic | 32 | 1M | - | 69 | 11.3 | anthropic |
| K2 Horizon 375B A23B | Institute of Foundation Models | 31 | 524k | - | - | - | — |
| Apodex 1.1 | Apodex | 30 | 256k | - | - | - | — |
| GPT-5.6 Terra (medium) | OpenAI | 30 | 1M | - | 83 | 2.0 | openai |
| Gemini 3.1 Pro Preview | 30 | 1M | - | 116 | 28.4 | ||
| GPT-6 Luna (medium) | OpenAI | 29 | 1M | - | 143 | 5.3 | openai |
| MiniMax-M3 | MiniMax | 29 | 1M | - | 118 | 1.1 | minimax |
| Nex-N2-Pro | Nex AGI | 28 | 262k | - | - | - | — |
| Solar Pro 4 | Upstage | 28 | 512k | - | 63 | 2.0 | upstage |
| GPT-6 Sol (Non-reasoning) | OpenAI | 28 | 872k | - | 109 | 0.9 | openai |
| Claude Sonnet 5 (medium) | Anthropic | 28 | 1M | - | 66 | 2.2 | anthropic |
| Inkling Small | Thinking Machines | 28 | 1M | - | 199 | 2.1 | — |
| Qwen3.8 27B (medium) | Alibaba | 28 | 256k | - | 49 | 3.9 | empiriolabs |
| GPT-5.6 Terra (low) | OpenAI | 27 | 1M | - | 82 | 1.8 | openai |
| JT-4.1 Flash 236B A21B | China Mobile | 27 | 256k | - | - | - | — |
| Agnes 2.5 Pro Alpha | Sapiens AI | 27 | 1M | - | 161 | 4.3 | agnes |
| Quasar 438B (max) | Multiverse Computing | 27 | 1M | - | 148 | 1.3 | — |
| Qwen3.8 27B (low) | Alibaba | 26 | 256k | - | 51 | 3.9 | empiriolabs |
| GPT-5.5 Instant (June 2026) | OpenAI | 26 | 400k | - | 129 | 1.1 | — |
| MiMo-V2.5-Pro | Xiaomi | 26 | 1M | - | 45 | 5.8 | xiaomi |
| Kimi K2.7 Code | Kimi | 26 | 256k | - | 61 | 3.3 | moonshotai |
| K2 Horizon MoVA 36B A4B | Institute of Foundation Models | 25 | 524k | - | - | - | — |
| Hy3 | Tencent | 25 | 256k | - | 87 | 2.9 | tencent-tokenhub |
| MiMo-V2.5 | Xiaomi | 25 | 1M | - | 35 | 3.7 | — |
| Qwen3.7 Plus | Alibaba | 25 | 1M | - | 65 | 2.2 | alibaba |
| Inkling | Thinking Machines | 25 | 1M | - | 114 | 2.3 | venice |
| Ling 3.0 Flash | InclusionAI | 25 | 262k | - | 329 | 2.5 | llmgateway |
| Solar Open2 250B | Upstage | 25 | 1.05M | - | - | - | — |
| Ling-3.0-flash-VL | InclusionAI | 25 | 262k | - | 146 | 2.8 | llmgateway |
| Claude Sonnet 5 (low) | Anthropic | 24 | 1M | - | 65 | 1.8 | anthropic |
| Claude Sonnet 5 (Non-reasoning) | Anthropic | 23 | 1M | - | 64 | 1.1 | anthropic |
| Nemotron 3 Ultra | NVIDIA | 23 | 262k | - | 162 | 2.3 | venice |
| A.X-K2 | SK Telecom | 23 | 262k | - | - | - | — |
| Ling-3.0-flash-Fin | InclusionAI | 23 | 262k | - | 159 | 2.6 | — |
| MiMo-V2-Flash (Feb 2026) | Xiaomi | 22 | 256k | - | - | - | — |
| Gemini 3.5 Flash-Lite | 22 | 1M | - | 361 | 8.4 | ||
| G9v3-39A5B | AI9Stars | 22 | 131k | - | - | - | — |
| KAT-Coder-Pro V2 | KwaiKAT | 22 | 256k | - | - | - | — |
| Qwen3.5 397B A17B (Non-reasoning) | Alibaba | 21 | 262k | - | 86 | 2.2 | alibaba |
| GPT-6 Luna (low) | OpenAI | 21 | 1M | - | 176 | 1.6 | openai |
| GPT-5.6 Terra (Non-reasoning) | OpenAI | 21 | 1M | - | 78 | 0.9 | openai |
| K2 Horizon 7B | Institute of Foundation Models | 21 | 524k | - | - | - | — |
| Qwen3.5 Omni Plus | Alibaba | 20 | 256k | - | 88 | 2.1 | — |
| o3 | OpenAI | 20 | 200k | - | 120 | 7.8 | openai |
| Qwen3.8 27B | Alibaba | 20 | 256k | - | 49 | 3.9 | empiriolabs |
| K-EXAONE 2.0 | LG AI Research | 20 | 262k | - | - | - | — |
| Step 3.7 Flash | StepFun | 19 | 262k | - | 194 | 2.6 | stepfun |
| LongCat 2.0 | LongCat | 19 | 1M | - | - | - | nano-gpt |
| Gemma 4 31B | 19 | 256k | - | 35 | 1.0 | neuralwatt | |
| JT-35B-Flash | China Mobile | 19 | 256k | - | - | - | — |
| Qwen3.5 397B A17B | Alibaba | 18 | 262k | - | 86 | 2.3 | alibaba |
| MiMo-V2.5-Pro (Non-reasoning) | Xiaomi | 18 | 1M | - | 45 | 5.0 | xiaomi |
| GPT-6 Luna (Non-reasoning) | OpenAI | 18 | 1M | - | 136 | 0.7 | openai |
| Qwen3.6 35B A3B | Alibaba | 18 | 262k | - | 116 | 2.2 | alibaba |
| Qwen3.5 122B A10B (Non-reasoning) | Alibaba | 18 | 262k | - | 142 | 2.4 | alibaba |
| Muse Glimmer (high) | Meta | 17 | 131k | - | 96 | 0.9 | — |
| Doubao Seed Code | ByteDance Seed | 17 | 256k | - | - | - | — |
| Claude 4.5 Haiku | Anthropic | 17 | 200k | - | 109 | 23.1 | — |
| Gemma 4 26B A4B | 17 | 256k | - | - | - | empiriolabs | |
| Ring-2.6-1T | InclusionAI | 17 | 262k | - | 119 | 3.7 | requesty |
| K2 Horizon 3.7B | Institute of Foundation Models | 16 | 524k | - | - | - | — |
| Qwen3.5 122B A10B | Alibaba | 16 | 262k | - | 127 | 2.4 | alibaba |
| Claude 4.5 Haiku (Non-reasoning) | Anthropic | 15 | 200k | - | 87 | 0.6 | helicone |
| Ling 3.0 Tiny | InclusionAI | 15 | 262k | - | 33 | 3.3 | requesty |
| Qwen3.6 35B A3B (Non-reasoning) | Alibaba | 15 | 262k | - | 132 | 2.2 | alibaba |
| Granite 4.2 30B | IBM | 15 | 131k | - | 70 | 0.9 | — |
| ERNIE 5.0 Thinking Preview | Baidu | 14 | 128k | - | - | - | nano-gpt |
| Mistral Medium 3.5 | Mistral | 14 | 256k | - | 146 | 2.3 | pioneer |
| Gemma 4 12B | 14 | 256k | - | 113 | 2.4 | — | |
| Nova 2.0 Pro Preview (medium) | Amazon | 14 | 256k | - | 120 | 14.5 | — |
| Gemma 4 31B (Non-reasoning) | 14 | 256k | - | 34 | 2.4 | neuralwatt | |
| Mercury 2 | Inception | 14 | 128k | - | 740 | 4.7 | inception |
| Qwen3.5 9B | Alibaba | 14 | 262k | - | 57 | 2.0 | empiriolabs |
| Nova 2.0 Omni (medium) | Amazon | 14 | 1M | - | - | - | — |
| Apriel-v1.6-15B-Thinker | ServiceNow | 13 | 128k | - | - | - | — |
| Nova 2.0 Lite (high) | Amazon | 13 | 1M | - | 166 | 15.2 | — |
| Qwen3.5 9B (Non-reasoning) | Alibaba | 13 | 262k | - | 88 | 0.8 | empiriolabs |
| EXAONE 4.5 33B | LG AI Research | 13 | 262k | - | - | - | — |
| Command A+ | Cohere | 13 | 192k | - | 240 | 0.4 | — |
| Gemma 4 26B A4B (Non-reasoning) | 13 | 256k | - | 67 | 1.3 | empiriolabs | |
| Qwen3.5 4B | Alibaba | 13 | 262k | - | 23 | 0.8 | empiriolabs |
| Nemotron 3.5 Lightning | NVIDIA | 13 | 1M | - | 277 | 0.6 | nano-gpt |
| Nemotron 3 Super | NVIDIA | 13 | 1M | - | 163 | 1.5 | requesty |
| Nova 2.0 Pro Preview (low) | Amazon | 13 | 256k | - | 115 | 10.8 | — |
| Nova 2.0 Lite (medium) | Amazon | 12 | 1M | - | 161 | 14.3 | — |
| Qwen3.5 Omni Flash | Alibaba | 12 | 256k | - | 226 | 2.0 | — |
| MiniCPM5-2B | OpenBMB | 12 | 131k | - | - | - | amd |
| JT-MINI | China Mobile | 12 | 128k | - | - | - | — |
| Magistral Medium 1.2 | Mistral | 12 | 128k | - | - | - | — |
| Nova 2.0 Lite (low) | Amazon | 12 | 1M | - | 163 | 8.9 | — |
| HyperNova 60B 2605 (high) | Multiverse Computing | 12 | 131k | - | - | - | — |
| Nemotron Cascade 2 30B A3B | NVIDIA | 12 | 1M | - | - | - | — |
| gpt-oss-120b (high) | OpenAI | 12 | 131k | - | 198 | 0.8 | cerebras |
| K2 Think V2 | Institute of Foundation Models | 11 | 262k | - | - | - | — |
| LongCat Flash Lite | LongCat | 11 | 256k | - | - | - | — |
| HyperCLOVA X SEED Think (32B) | Naver | 11 | 128k | - | - | - | — |
| Mistral Small 4 | Mistral | 11 | 256k | - | 170 | 0.7 | empiriolabs |
| Qwen3 Next 80B A3B (Reasoning) | Alibaba | 11 | 262k | - | 170 | 2.3 | — |
| Nova 2.0 Omni (low) | Amazon | 11 | 1M | - | - | - | — |
| Granite 4.2 8B | IBM | 11 | 131k | - | 97 | 0.9 | llmgateway |
| Mi:dm K 2.5 Pro | Korea Telecom | 11 | 128k | - | - | - | — |
| G9v3-3B | AI9Stars | 11 | 131k | - | - | - | — |
| Trinity Large Thinking | Arcee AI | 11 | 512k | - | 321 | 1.2 | arcee |
| Qwen3.5 4B (Non-reasoning) | Alibaba | 11 | 262k | - | 16 | 0.8 | empiriolabs |
| INTELLECT-3 | Prime Intellect | 11 | 131k | - | - | - | — |
| Solar Open 100B | Upstage | 10 | 128k | - | - | - | — |
| Nemotron 3 Nano Omni 30B A3B | NVIDIA | 10 | 256k | - | 269 | 0.5 | — |
| gpt-oss-120b (low) | OpenAI | 10 | 131k | - | 186 | 0.8 | cerebras |
| Llama 4 Maverick | Meta | 10 | 1M | - | 111 | 0.9 | neon |
| Nova 2.0 Pro Preview (Non-reasoning) | Amazon | 10 | 256k | - | 106 | 1.0 | — |
| gpt-oss-20b (low) | OpenAI | 10 | 131k | - | 224 | 0.9 | regolo-ai |
| North Mini Code | Cohere | 10 | 256k | - | 98 | 0.4 | — |
| K2-V2 (high) | Institute of Foundation Models | 10 | 512k | - | - | - | — |
| Qwen3 Next 80B A3B | Alibaba | 10 | 262k | - | 171 | 2.2 | alibaba |
| DiffusionGemma 26B A4B | 10 | 256k | - | - | - | — | |
| Gemma 4 12B (Non-reasoning) | 9 | 262k | - | 110 | 2.3 | — | |
| Mistral Large 3 | Mistral | 9 | 256k | - | 75 | 1.1 | pioneer |
| Qwen3 Coder Next | Alibaba | 9 | 256k | - | 91 | 1.6 | alibaba-coding-plan |
| Motif-2-12.7B | Motif Technologies | 9 | 128k | - | - | - | — |
| Nova Premier | Amazon | 9 | 1M | - | - | - | — |
| Granite 4.2 3B | IBM | 9 | 131k | - | 221 | 0.5 | — |
| K2-V2 (medium) | Institute of Foundation Models | 9 | 512k | - | - | - | — |
| Llama Nemotron Super 49B v1.5 | NVIDIA | 9 | 128k | - | 76 | 5.5 | — |
| Mistral Small 4 (Non-reasoning) | Mistral | 9 | 256k | - | 153 | 0.8 | empiriolabs |
| Tri-21B-Think | Trillion Labs | 9 | 32k | - | - | - | — |
| gpt-oss-20b (high) | OpenAI | 9 | 131k | - | 189 | 0.8 | regolo-ai |
| Gemma 4 E4B | 9 | 128k | - | 46 | 0.9 | — | |
| Nemotron 3 Nano | NVIDIA | 9 | 1M | - | 129 | 1.2 | — |
| MiniCPM5-1B | OpenBMB | 9 | 128k | - | - | - | — |
| Sarvam 105B (high) | Sarvam | 9 | 128k | - | - | - | sarvam |
| Nova 2.0 Lite (Non-reasoning) | Amazon | 9 | 1M | - | 164 | 1.0 | — |
| MiniCPM5-1B (Non-reasoning) | OpenBMB | 9 | 128k | - | - | - | — |
| Devstral 2 | Mistral | 9 | 256k | - | 136 | 2.3 | pioneer |
| Magistral Small 1.2 | Mistral | 9 | 128k | - | - | - | — |
| Nanbeige4.1-3B | Nanbeige | 8 | 256k | - | - | - | — |
| LFM2.5-2.6B | Liquid AI | 8 | 128k | - | - | - | — |
| EXAONE 4.0 32B | LG AI Research | 8 | 131k | - | - | - | — |
| Nova 2.0 Omni (Non-reasoning) | Amazon | 8 | 1M | - | - | - | — |
| Llama 4 Scout | Meta | 8 | 10M | - | 104 | 0.9 | helicone |
| Hermes 4 70B | Nous Research | 8 | 128k | - | - | - | — |
| Falcon-H1R-7B | TII UAE | 8 | 256k | - | - | - | — |
| Gemma 4 E2B | 8 | 128k | - | - | - | — | |
| Qwen3 Omni 30B A3B (Reasoning) | Alibaba | 8 | 65.5k | - | 102 | 2.0 | — |
| Step3 VL 10B | StepFun | 8 | 65.5k | - | - | - | — |
| Llama 3.3 70B | Meta | 8 | 128k | - | 83 | 1.5 | — |
| Devstral Small 2 | Mistral | 8 | 256k | - | 136 | 2.3 | pioneer |
| Llama Nemotron Ultra | NVIDIA | 8 | 128k | - | - | - | — |
| ERNIE 4.5 300B A47B | Baidu | 8 | 131k | - | - | - | — |
| Hermes 4 405B | Nous Research | 7 | 128k | - | 40 | 2.5 | — |
| NVIDIA Nemotron Nano 12B v2 VL | NVIDIA | 7 | 128k | - | 90 | 8.8 | — |
| Gemma 4 E4B (Non-reasoning) | 7 | 128k | - | 47 | 0.8 | — | |
| NVIDIA Nemotron Nano 9B V2 | NVIDIA | 7 | 131k | - | 113 | 9.9 | — |
| Hermes 4 405B (Non-reasoning) | Nous Research | 7 | 128k | - | 41 | 2.4 | — |
| Nemotron 3 Nano 4B | NVIDIA | 7 | 262k | - | - | - | — |
| Llama Nemotron Super 49B v1.5 (Non-reasoning) | NVIDIA | 7 | 128k | - | 62 | 5.8 | — |
| K2-V2 (low) | Institute of Foundation Models | 7 | 512k | - | - | - | — |
| Kimi Linear 48B A3B Instruct | Kimi | 7 | 1M | - | - | - | — |
| Llama 3.1 405B | Meta | 7 | 128k | - | - | - | — |
| LFM2.5-8B-A1B | Liquid AI | 7 | 32.8k | - | - | - | — |
| Ring-flash-2.0 | InclusionAI | 7 | 128k | - | - | - | — |
| Olmo 3.1 32B Think | Allen Institute for AI | 7 | 65.5k | - | - | - | — |
| Command A | Cohere | 7 | 256k | - | 61 | 1.8 | — |
| Qwen3.5 2B | Alibaba | 7 | 262k | - | - | - | requesty |
| Llama 3.1 Nemotron 70B | NVIDIA | 7 | 128k | - | 58 | 5.8 | — |
| Nemotron 3 Nano (Non-reasoning) | NVIDIA | 7 | 1M | - | 172 | 0.9 | venice |
| NVIDIA Nemotron Nano 9B V2 (Non-reasoning) | NVIDIA | 7 | 131k | - | 155 | 1.7 | vercel |
| Hermes 4 70B (Non-reasoning) | Nous Research | 7 | 128k | - | - | - | — |
| Sarvam 30B (high) | Sarvam | 7 | 65.5k | - | - | - | sarvam |
| Olmo 3.1 32B Instruct | Allen Institute for AI | 6 | 65.5k | - | - | - | — |
| Gemma 4 E2B (Non-reasoning) | 6 | 128k | - | - | - | — | |
| R1 1776 | Perplexity | 6 | 128k | - | - | - | — |
| Llama 3.2 90B (Vision) | Meta | 6 | 128k | - | - | - | — |
| Celeris-1 | Celeris | 6 | 131k | - | 1492 | 0.6 | nano-gpt |
| Phi-4 Mini | Microsoft | 6 | 128k | - | 45 | 0.9 | azure |
| EXAONE 4.0 32B (Non-reasoning) | LG AI Research | 6 | 131k | - | - | - | — |
| Qwen3.5 2B (Non-reasoning) | Alibaba | 6 | 262k | - | - | - | requesty |
| Qwen3.5 0.8B | Alibaba | 6 | 262k | - | - | - | qvac |
| DeepHermes 3 - Mistral 24B | Nous Research | 6 | 32k | - | - | - | — |
| Jamba 1.7 Large | AI21 Labs | 6 | 256k | - | - | - | — |
| Granite 4.0 H Small | IBM | 6 | 128k | - | 24 | 21.2 | — |
| Ministral 3 14B | Mistral | 6 | 256k | - | 89 | 0.9 | — |
| Qwen3 Omni 30B A3B | Alibaba | 6 | 65.5k | - | 96 | 1.9 | — |
| LFM2 24B A2B | Liquid AI | 6 | 32.8k | - | - | - | — |
| Phi-4 | Microsoft | 6 | 16k | - | 37 | 2.8 | azure |
| Nova Micro | Amazon | 6 | 130k | - | 270 | 0.9 | — |
| NVIDIA Nemotron Nano 12B v2 VL (Non-reasoning) | NVIDIA | 6 | 128k | - | 192 | 2.3 | vercel |
| Phi-4 Multimodal | Microsoft | 6 | 128k | - | 17 | 0.8 | azure |
| MiniCPM-V 4.6 1.3B | OpenBMB | 6 | 262k | - | - | - | — |
| Jamba Reasoning 3B | AI21 Labs | 6 | 262k | - | - | - | — |
| Reka Flash 3 | Reka AI | 6 | 128k | - | - | - | — |
| Olmo 3 7B Think | Allen Institute for AI | 6 | 65.5k | - | - | - | — |
| Molmo 7B-D | Allen Institute for AI | 6 | 4.1k | - | - | - | — |
| Ministral 3 8B | Mistral | 5 | 256k | - | 82 | 0.8 | — |
| Llama 3.2 11B (Vision) | Meta | 5 | 128k | - | 14 | 2.0 | — |
| Qwen3.5 0.8B (Non-reasoning) | Alibaba | 5 | 262k | - | - | - | qvac |
| Exaone 4.0 1.2B | LG AI Research | 5 | 64k | - | - | - | — |
| Olmo 3 7B | Allen Institute for AI | 5 | 65.5k | - | - | - | — |
| Exaone 4.0 1.2B (Non-reasoning) | LG AI Research | 5 | 64k | - | - | - | — |
| LFM2.5-1.2B-Thinking | Liquid AI | 5 | 32k | - | - | - | — |
| Jamba 1.7 Mini | AI21 Labs | 5 | 258k | - | - | - | — |
| LFM2.5-1.2B-Instruct | Liquid AI | 5 | 32k | - | - | - | — |
| Granite 4.0 H 1B | IBM | 5 | 128k | - | - | - | — |
| Gemma 3 270M | 5 | 32k | - | - | - | — | |
| Apertus 70B Instruct | Swiss AI Initiative | 5 | 65.5k | - | - | - | — |
| Granite 4.0 Micro | IBM | 5 | 128k | - | - | - | — |
| DeepHermes 3 - Llama-3.1 8B | Nous Research | 5 | 128k | - | - | - | — |
| Molmo2-8B | Allen Institute for AI | 5 | 36.9k | - | - | - | — |
| Ministral 3 3B | Mistral | 5 | 256k | - | 224 | 0.7 | — |
| LFM2.5-VL-1.6B | Liquid AI | 5 | 32k | - | - | - | — |
| Granite 4.0 350M | IBM | 5 | 32.8k | - | - | - | — |
| Tiny Aya Global | Cohere | 5 | 8.19k | - | - | - | — |
| Apertus 8B Instruct | Swiss AI Initiative | 5 | 65.5k | - | - | - | — |
| Granite 4.0 H 350M | IBM | 5 | 32.8k | - | - | - | — |
| K2 Horizon 0.9B | Institute of Foundation Models | 3 | 131k | - | - | - | — |
| EXAONE 4.5 33B (Non-reasoning) | LG AI Research | - | 262k | - | - | - | — |
| Gemini 3 Deep Think | - | 128k | - | - | - | — | |
| GPT-5.5 Pro (xhigh) | OpenAI | - | 922k | - | - | - | openai |
| Cogito v2.1 | Deep Cogito | - | 128k | - | - | - | — |