Best LLM for agents
Agentic tool use covers whether a model drives tools to a finished outcome, stays steerable, and recovers when a command fails. Terminal-Bench and other evaluator-run agentic results identify the model together with its agent harness and effort; they are not pure model-only scores.
Aldena runs these models inside your team's projects. See what each one costs.
| rank | model | vendor | composite | pricein / out | benchmarks | % | % | score | score | score | score | score |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Claude Sonnet 5.5 (max)anthropic/claude-sonnet-5.5:max | Anthropic | 80.8 | $2.00 / $10.00 | 6 of 7 benchmarks | 63.9% | 16.0 | 13.8 | 19.5 | 0.8 | 15.0 | |
| 2 | Claude Opus 5.5 (max)anthropic/claude-opus-5.5:max | Anthropic | 79.6 | $4.00 / $20.00 | 5 of 7 benchmarks | 65.8% | 20.0 | 18.6 | 20.7 | 0.7 | ||
| 3 | GPT-6 Astra (max)openai/gpt-6-astra:max | OpenAI | 78.0 | $10.00 / $50.00 | 6 of 7 benchmarks | 60.3% | 14.2 | 14.8 | 13.6 | 0.9 | 5.5 | |
| 4 | Claude Fable 5.1 (max)anthropic/claude-fable-5.1:max | Anthropic | 77.2 | $10.00 / $50.00 | 6 of 7 benchmarks | 57.3% | 14.3 | 15.7 | 13.1 | 0.8 | 11.6 | |
| 5 | GPT-6.1 Terra (max)openai/gpt-6.1-terra:max | OpenAI | 76.5 | $1.00 / $5.00 | 4 of 7 benchmarks | 12.1 | 15.8 | 10.6 | 0.9 | |||
| 6 | GPT-6.1 Sol (max)openai/gpt-6.1-sol:max | OpenAI | 76.0 | $2.00 / $10.00 | 6 of 7 benchmarks | 60.9% | 12.5 | 13.3 | 11.2 | 0.9 | 4.5 | |
| 7 | Gemini 4 Carbon (high)google/gemini-4-carbon:high | 73.7 | — | 4 of 7 benchmarks | 16.6 | 19.1 | 18.4 | 0.4 | ||||
| 8 | Claude Opus 5 (max)anthropic/claude-opus-5:max | Anthropic | 71.9 | $5.00 / $25.00 | 6 of 7 benchmarks | 52.1% | 7.5 | 4.9 | 10.6 | 0.9 | 13.6 | |
| 9 | Claude Fable 5 (high)anthropic/claude-fable-5:high | Anthropic | 71.0 | $10.00 / $50.00 | 5 of 7 benchmarks | 7.7 | 8.5 | 5.7 | 0.7 | 9.6 | ||
| 10 | GPT-6 Sol (max)openai/gpt-6-sol:max | OpenAI | 69.9 | $2.00 / $10.00 | 6 of 7 benchmarks | 50.3% | 7.0 | 10.1 | 3.8 | 0.9 | 7.0 | |
| 11 | Kimi K3.1 Pro (max)moonshotai/kimi-k3.1-pro:max | MoonshotAI | 68.1 | $2.00 / $10.00 | 4 of 7 benchmarks | 5.2 | 7.5 | 2.8 | 0.9 | |||
| 12 | Gemini 4 Argongoogle/gemini-4-argon | 66.5 | $2.00 / $10.00 | 1 of 7 benchmarks | 90.7% | |||||||
| 13 | Claude Opus 5.5 (high)anthropic/claude-opus-5.5:high | Anthropic | 65.2 | $4.00 / $20.00 | 1 of 7 benchmarks | 13.6 | ||||||
| 14 | Claude Opus 5 (xhigh)anthropic/claude-opus-5:xhigh | Anthropic | 64.7 | $5.00 / $25.00 | 2 of 7 benchmarks | 53.3% | 85.8% | |||||
| 15 | Claude Fable 5.1anthropic/claude-fable-5.1 | Anthropic | 64.7 | $10.00 / $50.00 | 1 of 7 benchmarks | 87.2% | ||||||
| 16 | Hy4tencent/hy4 | Tencent | 63.2 | $0.417 / $1.2505 | 4 of 7 benchmarks | 4.7 | 5.0 | 8.1 | 0.1 | |||
| 17 | Muse Spark 1.3 (max)meta/muse-spark-1.3:max | Meta | 63.0 | $1.25 / $4.25 | 5 of 7 benchmarks | 2.2 | 2.9 | 3.7 | 0.7 | 6.1 | ||
| 18 | GPT-5.6 Sol (xhigh)openai/gpt-5.6-sol:xhigh | OpenAI | 62.7 | $4.00 / $20.00 | 5 of 7 benchmarks | 4.5 | 6.1 | 1.9 | 0.9 | 0.7 | ||
| 19 | GPT-6 Astra (high)openai/gpt-6-astra:high | OpenAI | 62.6 | $10.00 / $50.00 | 1 of 7 benchmarks | 60.0% | ||||||
| 20 | Grok 4.7 (xhigh)x-ai/grok-4.7:xhigh | SpaceXAI | 62.6 | $2.00 / $6.00 | 6 of 7 benchmarks | 38.8% | 2.6 | -2.0 | 5.5 | 0.9 | 7.2 | |
| 21 | Claude Opus 5 (high)anthropic/claude-opus-5:high | Anthropic | 62.4 | $5.00 / $25.00 | 2 of 7 benchmarks | 49.7% | 12.6 | |||||
| 22 | Claude Opus 5.5anthropic/claude-opus-5.5 | Anthropic | 62.4 | $4.00 / $20.00 | 1 of 7 benchmarks | 85.4% | ||||||
| 23 | GPT-6 Astra (xhigh)openai/gpt-6-astra:xhigh | OpenAI | 61.6 | $10.00 / $50.00 | 1 of 7 benchmarks | 59.7% | ||||||
| 24 | DeepSeek V4.1 Flash (max)deepseek/deepseek-v4.1-flash:max | DeepSeek | 61.5 | $0.051 / $0.567 | 5 of 7 benchmarks | 2.1 | 3.5 | 2.5 | 0.3 | 7.9 | ||
| 25 | Kimi K3 (max)moonshotai/kimi-k3:max | MoonshotAI | 61.5 | $0.192 / $16.00 | 6 of 7 benchmarks | 82.3% | 0.6 | -0.4 | 3.5 | 0.9 | 4.4 | |
| 26 | Qwen3.8 2.4T A95B (xhigh)qwen/qwen3.8-2.4t-a95b:xhigh | Qwen | 61.0 | $2.00 / $6.00 | 1 of 7 benchmarks | 84.5% | ||||||
| 27 | Claude Fable 5.1 (xhigh)anthropic/claude-fable-5.1:xhigh | Anthropic | 60.6 | $10.00 / $50.00 | 1 of 7 benchmarks | 57.6% | ||||||
| 28 | Claude Opus 4.8 (high)anthropic/claude-opus-4.8:high | Anthropic | 60.6 | $5.00 / $25.00 | 1 of 7 benchmarks | 7.2 | ||||||
| 29 | Gemini 4 Argon (high)google/gemini-4-argon:high | 60.6 | $2.00 / $10.00 | 5 of 7 benchmarks | 10.8 | 13.5 | 14.1 | 0.4 | -21.4 | |||
| 30 | GLM 5.3z-ai/glm-5.3 | Z.ai | 60.0 | $0.0699 / $1.40 | 1 of 7 benchmarks | 84.2% | ||||||
| 31 | Claude Sonnet 5 (high)anthropic/claude-sonnet-5:high | Anthropic | 59.3 | $2.00 / $10.00 | 5 of 7 benchmarks | 2.1 | 3.0 | -0.5 | 0.5 | 7.4 | ||
| 32 | Gemini 3.5 Flash (high)google/gemini-3.5-flash:high | 59.1 | $1.50 / $9.00 | 1 of 7 benchmarks | 83.6% | |||||||
| 33 | GPT-6 Astra (medium)openai/gpt-6-astra:medium | OpenAI | 58.7 | $10.00 / $50.00 | 1 of 7 benchmarks | 55.1% | ||||||
| 34 | Claude Fable 5anthropic/claude-fable-5 | Anthropic | 58.2 | $10.00 / $50.00 | 1 of 7 benchmarks | 83.3% | ||||||
| 35 | Claude Fable 5.1 (high)anthropic/claude-fable-5.1:high | Anthropic | 57.7 | $10.00 / $50.00 | 1 of 7 benchmarks | 54.9% | ||||||
| 36 | GPT-5.5 (xhigh)openai/gpt-5.5:xhigh | OpenAI | 57.4 | $5.00 / $30.00 | 2 of 7 benchmarks | 75.3% | 9.6 | |||||
| 37 | Claude Fable 5.1 (medium)anthropic/claude-fable-5.1:medium | Anthropic | 56.7 | $10.00 / $50.00 | 1 of 7 benchmarks | 53.9% | ||||||
| 38 | MiMo-V2.6-Proxiaomi/mimo-v2.6-pro | Xiaomi | 56.2 | $0.435 / $0.87 | 5 of 7 benchmarks | 2.3 | 0.8 | 4.9 | 0.0 | 2.3 | ||
| 39 | Muse Sparkmeta/muse-spark | Meta | 55.9 | — | 1 of 7 benchmarks | 82.2% | ||||||
| 40 | Hy4 previewtencent/hy4-preview | Tencent | 55.8 | $0.7506 / $2.2509 | 5 of 7 benchmarks | 0.4 | -0.4 | 4.2 | -0.0 | 6.5 | ||
| 41 | Muse Wonder (max)meta/muse-wonder:max | Meta | 55.8 | $4.00 / $20.00 | 4 of 7 benchmarks | 1.9 | 2.6 | 5.4 | -0.8 | |||
| 42 | GPT-6 Luna (max)openai/gpt-6-luna:max | OpenAI | 54.9 | $0.10 / $0.50 | 6 of 7 benchmarks | 15.2% | 1.5 | 3.8 | -0.6 | 0.9 | 5.2 | |
| 43 | GLM 5.2 (max)z-ai/glm-5.2:max | Z.ai | 54.5 | $0.006 / $6.80 | 5 of 7 benchmarks | 0.4 | -0.9 | 2.0 | 0.9 | 0.7 | ||
| 44 | GPT-5.6 Solopenai/gpt-5.6-sol | OpenAI | 54.5 | $4.00 / $20.00 | 1 of 7 benchmarks | 81.8% | ||||||
| 45 | Grok 4.6 (xhigh)x-ai/grok-4.6:xhigh | SpaceXAI | 54.5 | $2.00 / $6.00 | 5 of 7 benchmarks | -0.2 | -1.0 | -2.5 | 0.9 | 5.7 | ||
| 46 | Grok 4.5x-ai/grok-4.5 | SpaceXAI | 54.1 | $2.00 / $6.00 | 1 of 7 benchmarks | 5.1 | ||||||
| 47 | GPT-6 Astra (low)openai/gpt-6-astra:low | OpenAI | 53.8 | $10.00 / $50.00 | 1 of 7 benchmarks | 50.9% | ||||||
| 48 | GLM 5.3 (max)z-ai/glm-5.3:max | Z.ai | 53.6 | $0.0699 / $1.40 | 6 of 7 benchmarks | 41.2% | 1.0 | -0.9 | 2.9 | 0.9 | -0.1 | |
| 49 | Inkling 2 Smallthinkingmachines/inkling-2-small | Thinking Machines | 53.2 | — | 5 of 7 benchmarks | 85.4% | 0.1 | 2.1 | 2.1 | -1.5 | ||
| 50 | DeepSeek V4 Pro 0423 (high)deepseek/deepseek-v4-pro:high | DeepSeek | 52.8 | $0.8837 / $1.7675 | 1 of 7 benchmarks | 4.4 | ||||||
| 51 | Claude Opus 4.7 (max)anthropic/claude-opus-4.7:max | Anthropic | 52.6 | $5.00 / $25.00 | 1 of 7 benchmarks | 79.1% | ||||||
| 52 | GPT-6 Luna (medium)openai/gpt-6-luna:medium | OpenAI | 51.8 | $0.10 / $0.50 | 4 of 7 benchmarks | -2.0 | 0.1 | -6.3 | 0.9 | |||
| 53 | Gemini 3.1 Pro Preview (high)google/gemini-3.1-pro-preview:high | 51.7 | $2.00 / $12.00 | 1 of 7 benchmarks | 78.2% | |||||||
| 54 | GPT-5.5openai/gpt-5.5 | OpenAI | 51.5 | $5.00 / $30.00 | 1 of 7 benchmarks | 4.1 | ||||||
| 55 | Claude Opus 5 (medium)anthropic/claude-opus-5:medium | Anthropic | 50.8 | $5.00 / $25.00 | 1 of 7 benchmarks | 45.1% | ||||||
| 56 | GLM 5.2z-ai/glm-5.2 | Z.ai | 50.8 | $0.006 / $6.80 | 1 of 7 benchmarks | 77.8% | ||||||
| 57 | Qwen3.8 Maxqwen/qwen3.8-max | Qwen | 50.1 | — | 5 of 7 benchmarks | 0.3 | -0.7 | 4.7 | -0.8 | 2.8 | ||
| 58 | Claude Fable 5 (max)anthropic/claude-fable-5:max | Anthropic | 49.9 | $10.00 / $50.00 | 1 of 7 benchmarks | 44.9% | ||||||
| 59 | Claude Opus 4.6 (max)anthropic/claude-opus-4.6:max | Anthropic | 49.9 | $5.00 / $25.00 | 1 of 7 benchmarks | 76.8% | ||||||
| 60 | Inkling (xhigh)thinkingmachines/inkling:xhigh | Thinking Machines | 48.9 | $1.00 / $4.05 | 1 of 7 benchmarks | 76.0% | ||||||
| 61 | Claude Fable 5.1 (low)anthropic/claude-fable-5.1:low | Anthropic | 48.9 | $10.00 / $50.00 | 1 of 7 benchmarks | 43.3% | ||||||
| 62 | GPT-5.4 (high)openai/gpt-5.4:high | OpenAI | 48.9 | $2.50 / $15.00 | 1 of 7 benchmarks | 1.2 | ||||||
| 63 | GPT-5.6 Terra (xhigh)openai/gpt-5.6-terra:xhigh | OpenAI | 48.8 | $2.00 / $12.00 | 5 of 7 benchmarks | -1.9 | 0.2 | -7.5 | 0.9 | 0.9 | ||
| 64 | Claude Opus 4.8 (max)anthropic/claude-opus-4.8:max | Anthropic | 48.5 | $5.00 / $25.00 | 2 of 7 benchmarks | 23.6% | 82.2% | |||||
| 65 | GLM 5.1z-ai/glm-5.1 | Z.ai | 48.0 | $1.40 / $4.40 | 1 of 7 benchmarks | 75.6% | ||||||
| 66 | Claude Haiku 5.5 (high)anthropic/claude-haiku-5.5:high | Anthropic | 47.8 | $0.10 / $0.50 | 4 of 7 benchmarks | 0.4 | 2.7 | 1.3 | -1.7 | |||
| 67 | GPT-5.4 (xhigh)openai/gpt-5.4:xhigh | OpenAI | 46.2 | $2.50 / $15.00 | 1 of 7 benchmarks | 70.6% | ||||||
| 68 | Gemini 3.8 Flash (high)google/gemini-3.8-flash:high | 46.1 | $0.75 / $3.75 | 6 of 7 benchmarks | 18.2% | 2.4 | -0.6 | 0.9 | 0.5 | -0.7 | ||
| 69 | GPT-5.6 Sol (max)openai/gpt-5.6-sol:max | OpenAI | 45.9 | $4.00 / $20.00 | 1 of 7 benchmarks | 38.5% | ||||||
| 70 | GLM 5.3 Flashz-ai/glm-5.3-flash | Z.ai | 45.8 | $0.15 / $0.50 | 5 of 7 benchmarks | -2.8 | -4.7 | 1.7 | 0.9 | -5.7 | ||
| 71 | MiMo-V2.6-Flashxiaomi/mimo-v2.6-flash | Xiaomi | 45.7 | $0.14 / $0.28 | 5 of 7 benchmarks | 0.1 | 2.3 | 5.0 | -4.0 | -1.1 | ||
| 72 | GPT-5.6 Luna (xhigh)openai/gpt-5.6-luna:xhigh | OpenAI | 45.6 | $0.20 / $1.20 | 1 of 7 benchmarks | 0.5 | ||||||
| 73 | Gemini 3 Pro Previewgoogle/gemini-3-pro-preview | 45.2 | — | 1 of 7 benchmarks | 70.3% | |||||||
| 74 | GLM 5.3 Flash (none)z-ai/glm-5.3-flash:none | Z.ai | 45.0 | $0.15 / $0.50 | 1 of 7 benchmarks | 35.5% | ||||||
| 75 | Claude Opus 4.5 (high)anthropic/claude-opus-4.5:high | Anthropic | 44.3 | $5.00 / $25.00 | 1 of 7 benchmarks | 69.8% | ||||||
| 76 | Claude Opus 5 (low)anthropic/claude-opus-5:low | Anthropic | 44.0 | $5.00 / $25.00 | 1 of 7 benchmarks | 34.9% | ||||||
| 77 | Claude Sonnet 4.6anthropic/claude-sonnet-4.6 | Anthropic | 43.4 | $3.00 / $15.00 | 1 of 7 benchmarks | 69.5% | ||||||
| 78 | Muse Spark 1.1meta/muse-spark-1.1 | Meta | 43.3 | $1.25 / $4.25 | 6 of 7 benchmarks | 88.1% | -7.9 | -14.6 | -6.5 | 0.9 | -2.3 | |
| 79 | MiniMax M3.1 Flashminimax/minimax-m3.1-flash | MiniMax | 43.1 | $0.075 / $0.30 | 4 of 7 benchmarks | -2.8 | -3.5 | 0.3 | 0.2 | |||
| 80 | Qwen3.8 Max (0902) (max)qwen/qwen3.8-max-0902:max | Qwen | 43.0 | $2.00 / $6.00 | 1 of 7 benchmarks | 27.0% | ||||||
| 81 | GPT-6 Luna (low)openai/gpt-6-luna:low | OpenAI | 42.8 | $0.10 / $0.50 | 4 of 7 benchmarks | -6.2 | -3.3 | -15.6 | 0.9 | |||
| 82 | GPT-5.2 (xhigh)openai/gpt-5.2:xhigh | OpenAI | 42.4 | $1.75 / $14.00 | 1 of 7 benchmarks | 67.6% | ||||||
| 83 | Kimi K2p5moonshotai/kimi-k2p5 | Moonshot AI | 41.5 | — | 1 of 7 benchmarks | 64.4% | ||||||
| 84 | GPT-5.6 Terra (max)openai/gpt-5.6-terra:max | OpenAI | 41.0 | $2.00 / $12.00 | 1 of 7 benchmarks | 21.5% | ||||||
| 85 | Nemotron 3 Ultra (thinking)nvidia/nemotron-3-ultra-550b-a55b:thinking | NVIDIA | 40.6 | $0.60 / $2.40 | 1 of 7 benchmarks | 63.1% | ||||||
| 86 | Muse Spark 1.2 (xhigh)meta/muse-spark-1.2:xhigh | Meta | 40.5 | $1.25 / $4.25 | 5 of 7 benchmarks | -8.0 | -11.8 | -9.3 | 0.7 | 5.8 | ||
| 87 | Grok 4.6 (high)x-ai/grok-4.6:high | SpaceXAI | 40.1 | $2.00 / $6.00 | 1 of 7 benchmarks | 19.7% | ||||||
| 88 | Gemini 3 Flash Previewgoogle/gemini-3-flash-preview | 39.7 | $0.50 / $3.00 | 1 of 7 benchmarks | 62.0% | |||||||
| 89 | Claude Sonnet 4.5 (thinking)anthropic/claude-sonnet-4.5:thinking | Anthropic | 38.7 | $3.00 / $15.00 | 1 of 7 benchmarks | 59.5% | ||||||
| 90 | GPT-5.6 Luna (max)openai/gpt-5.6-luna:max | OpenAI | 38.1 | $0.20 / $1.20 | 1 of 7 benchmarks | 17.6% | ||||||
| 91 | Step 5 Previewstepfun/step-5-preview | StepFun | 38.0 | $1.00 / $2.70 | 5 of 7 benchmarks | -3.4 | -1.1 | -0.7 | -1.6 | 0.2 | ||
| 92 | Mistral Large 4mistralai/mistral-large-4-0 | Mistral | 37.9 | $0.70 / $2.10 | 5 of 7 benchmarks | -6.1 | -6.9 | -5.3 | -0.0 | 1.8 | ||
| 93 | GLM 4.7z-ai/glm-4.7 | Z.ai | 37.8 | $0.60 / $2.20 | 1 of 7 benchmarks | 58.1% | ||||||
| 94 | Gemini 3.1 Flash Lite (high)google/gemini-3.1-flash-lite:high | 36.9 | $0.25 / $1.50 | 1 of 7 benchmarks | 57.1% | |||||||
| 95 | MiMo-V2.5-Proxiaomi/mimo-v2.5-pro | Xiaomi | 36.4 | $0.48 / $1.80 | 1 of 7 benchmarks | -7.5 | ||||||
| 96 | Qwen3.8 Flash Nextqwen/qwen3.8-flash-next | Qwen | 36.4 | — | 5 of 7 benchmarks | -4.7 | -2.9 | 0.0 | -1.0 | -2.6 | ||
| 97 | Muse Spark 1.3 (xhigh)meta/muse-spark-1.3:xhigh | Meta | 36.1 | $1.25 / $4.25 | 1 of 7 benchmarks | 14.8% | ||||||
| 98 | GPT-5.4 Mini (xhigh)openai/gpt-5.4-mini:xhigh | OpenAI | 36.0 | $0.75 / $4.50 | 1 of 7 benchmarks | 56.7% | ||||||
| 99 | Gemini 3.6 Flash (high)google/gemini-3.6-flash:high | 35.8 | $0.75 / $3.75 | 1 of 7 benchmarks | -18.2 | |||||||
| 100 | Qwen3.8 27Bqwen/qwen3.8-27b | Qwen | 35.8 | $0.15 / $1.875 | 5 of 7 benchmarks | -4.6 | -5.4 | -2.9 | 0.1 | -5.9 | ||
| 101 | GPT-5.1 (high)openai/gpt-5.1:high | OpenAI | 35.0 | $1.25 / $10.00 | 1 of 7 benchmarks | 50.1% | ||||||
| 102 | Gemini 3.7 Flash (high)google/gemini-3.7-flash:high | 35.0 | $0.75 / $3.75 | 6 of 7 benchmarks | 12.4% | -1.9 | -6.4 | -3.5 | 0.6 | -5.4 | ||
| 103 | Grok 4.5 (high)x-ai/grok-4.5:high | SpaceXAI | 34.7 | $2.00 / $6.00 | 1 of 7 benchmarks | 12.4% | ||||||
| 104 | o3 Proopenai/o3-pro | OpenAI | 34.1 | $20.00 / $80.00 | 1 of 7 benchmarks | 44.5% | ||||||
| 105 | Claude Haiku 4.5anthropic/claude-haiku-4.5 | Anthropic | 33.2 | $1.00 / $5.00 | 1 of 7 benchmarks | 40.2% | ||||||
| 106 | Claude Sonnet 5 (max)anthropic/claude-sonnet-5:max | Anthropic | 33.2 | $2.00 / $10.00 | 1 of 7 benchmarks | 12.1% | ||||||
| 107 | Qwen3.7 Maxqwen/qwen3.7-max | Qwen | 30.3 | $1.475 / $4.425 | 5 of 7 benchmarks | -7.1 | -11.8 | -10.7 | -0.6 | -1.6 | ||
| 108 | Inkling Smallthinkingmachines/inkling-small | Thinking Machines | 30.0 | $0.45 / $1.20 | 6 of 7 benchmarks | 79.2% | -17.7 | -26.7 | -27.3 | -3.6 | 5.4 | |
| 109 | Gemini 3.1 Pro Previewgoogle/gemini-3.1-pro-preview | 29.3 | $2.00 / $12.00 | 5 of 7 benchmarks | -4.3 | -7.7 | -11.5 | -0.2 | -23.3 | |||
| 110 | Beam 1.0 501B A23Breflection/beam-1.0-501b-a23b | Reflection | 29.3 | $0.30 / $1.00 | 4 of 7 benchmarks | -6.4 | -9.5 | -11.3 | -1.8 | |||
| 111 | MiniMax M3minimax/minimax-m3 | MiniMax | 29.1 | $0.30 / $1.20 | 5 of 7 benchmarks | -10.4 | -10.3 | -17.7 | -1.1 | 0.6 | ||
| 112 | Hy3tencent/hy3 | Tencent | 26.5 | $0.0825 / $0.33 | 5 of 7 benchmarks | -7.7 | -2.6 | -12.1 | -4.7 | -7.2 | ||
| 113 | Qwen3.7 Plusqwen/qwen3.7-plus | Qwen | 25.7 | $0.32 / $1.28 | 5 of 7 benchmarks | -7.9 | -11.3 | -10.4 | -1.9 | -6.3 | ||
| 114 | Nemotron 3 Ultranvidia/nemotron-3-ultra-550b-a55b | NVIDIA | 24.0 | $0.60 / $2.40 | 4 of 7 benchmarks | -14.4 | -15.4 | -22.2 | -1.4 | |||
| 115 | Inklingthinkingmachines/inkling | Thinking Machines | 21.9 | $1.00 / $4.05 | 5 of 7 benchmarks | -17.1 | -21.7 | -26.3 | -3.4 | 0.1 | ||
| 116 | Nemotron 3.5 Super VLnvidia/nemotron-3.5-super-vl | NVIDIA | 19.7 | $0.15 / $0.47 | 4 of 7 benchmarks | -16.3 | -17.8 | -18.9 | -7.0 | |||
| 117 | Mistral Medium 3.5mistralai/mistral-medium-3-5 | Mistral | 19.0 | $1.50 / $7.50 | 5 of 7 benchmarks | -13.9 | -16.6 | -17.7 | -6.1 | -20.4 | ||
| 118 | Solar Pro 4upstage/solar-pro4 | Upstage | 15.8 | $0.15 / $0.60 | 5 of 7 benchmarks | -22.8 | -20.1 | -31.3 | -3.7 | -25.6 |
How this ranks
Every benchmark value becomes a percentile among the models that have it, so accuracy scores, Elo ratings and word error rates compare without hand-tuned scaling. Metrics where lower is better are inverted first. Raw values are never summed or averaged across benchmarks. A model's mean percentile is then shrunk toward the mean of the models that were broadly benchmarked, so a model tested twice cannot outrank a broadly tested one on two lucky results. Turning a data source off runs that same ranking code again in your browser over the sources you left on.
A model scored on fewer than 3 of the 7 ranked benchmarks in this category still ranks here, on the benchmarks it does have, and its row carries a partial coverage mark. On an equal score it sits under the model that earned the same number across more of the board.
Data sources
Turn a source off to drop every benchmark it feeds and re-rank the board in your browser. Your choice follows you across the leaderboard pages.
- LMArenaCC BY 4.0
Arena ratings by LMArena, from the public leaderboard dataset.
- MCP AtlasMIT
All-1,000-task Pass Rate from Scale's current Performance Comparison, using the standardized MCP loop and 100-tool-call budget.
- Terminal-Bench 4.1Apache-2.0
Published Accuracy over 66 Terminal-Bench 4.1 tasks, run and verified by the benchmark team; results are model plus agent harness plus effort, not model-only evaluations.