Multi-step tool-using tasks judged by humans on real agentic sessions — the closest signal to "will this model run my agent well". Ranked by Arena agent, priced against live API rates. 28 models have a published score on this board.
Capability published 2026-10-23 · pricing retrieved 2026-10-25
anthropic/claude-opus-5.5anthropic/claude-opus-5.5anthropic/claude-sonnet-5.5anthropic/claude-opus-5.5Rows marked frontier are Pareto-optimal — nothing we track is both stronger and cheaper. "Strength" rescales the board so 100% = parity with the leader.
| # | Model | Provider | Score | Strength | $/1M | Context | Tools |
|---|---|---|---|---|---|---|---|
| 1 | Claude Opus 5.5anthropic/claude-opus-5.5 | Anthropic | 0.2077 | 100% | $8.00 | 1M | tools |
| 2 | Claude Sonnet 5.5anthropic/claude-sonnet-5.5 | Anthropic | 0.1399 | 67% | $4.00 | 1M | tools |
| 3 | GPT-6 Astraopenai/gpt-6-astra | OpenAI | 0.1396 | 67% | $20.00 | 1.05M | tools |
| 4 | Claude Fable 5.1anthropic/claude-fable-5.1 | Anthropic | 0.1394 | 67% | $20.00 | 1M | tools |
| 5 | GPT-6.1 Solopenai/gpt-6.1-sol | OpenAI | 0.1364 | 66% | $4.00 | 1.05M | tools |
| 6 | Gemini 4 Argongoogle/gemini-4-argon | 0.1175 | 57% | $4.00 | 1.048576M | tools | |
| 7 | GPT-6 Solopenai/gpt-6-sol | OpenAI | 0.0857 | 41% | $4.00 | 1.05M | tools |
| 8 | Claude Opus 5anthropic/claude-opus-5 | Anthropic | 0.0822 | 40% | $10.00 | 1M | tools |
| 9 | Claude Fable 5anthropic/claude-fable-5 | Anthropic | 0.0775 | 37% | $20.00 | 1M | tools |
| 10 | Claude Sonnet 5anthropic/claude-sonnet-5 | Anthropic | 0.0375 | 18% | $4.00 | 1M | tools |
| 11 | Gemini 3.8 Flashgoogle/gemini-3.8-flash | 0.0317 | 15% | $1.66 | 1.048576M | tools | |
| 12 | Grok 4.7x-ai/grok-4.7 | xAI | 0.0302 | 15% | $3.00 | 500K | tools |
| 13 | GPT-6 Lunaopenai/gpt-6-luna | OpenAI | 0.0279 | 13% | $0.20 | 1.05M | tools |
| 14 | DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash | DeepSeek | 0.0225 | 11% | $0.27 | 1.048576M | tools |
| 15 | GLM 5.2z-ai/glm-5.2 | Z.ai (Zhipu) | 0.0169 | 8% | $1.77 | 1.048576M | tools |
| 16 | Kimi K3moonshotai/kimi-k3 | Moonshot AI | 0.0157 | 8% | $2.88 | 1.048576M | tools |
| 17 | GLM 5.3z-ai/glm-5.3 | Z.ai (Zhipu) | 0.0122 | 6% | $0.41 | 1.048576M | tools |
| 18 | Claude Haiku 5.5anthropic/claude-haiku-5.5 | Anthropic | 0.0114 | 5% | $0.20 | 1M | tools |
| 19 | Qwen3.8 Max (0902)qwen/qwen3.8-max-0902 | Qwen (Alibaba) | 0.0113 | 5% | $3.00 | 1M | tools |
| 20 | Grok 4.6x-ai/grok-4.6 | xAI | 0.0063 | 3% | $3.00 | 500K | tools |
| 21 | DeepSeek V4 Pro 0813deepseek/deepseek-v4-pro-0813 | DeepSeek | -0.0027 | -1% | $0.99 | 1.048576M | tools |
| 22 | DeepSeek V4 Pro 0423deepseek/deepseek-v4-pro | DeepSeek | -0.0027 | -1% | $0.26 | 1.048576M | tools |
| 23 | Gemini 3.7 Flashgoogle/gemini-3.7-flash | -0.0127 | -6% | $1.50 | 1.048576M | tools | |
| 24 | GLM 5.3 Flashz-ai/glm-5.3-flash | Z.ai (Zhipu) | -0.0177 | -9% | $0.24 | 1.048576M | tools |
| 25 | Qwen3.8 27Bqwen/qwen3.8-27b | Qwen (Alibaba) | -0.0410 | -20% | $0.78 | 1M | tools |
Capability: Arena leaderboard dataset (CC-BY-4.0), Arena agent, published 2026-10-23. Pricing: OpenRouter, retrieved 2026-10-25. (input x 3 + output x 1) / 4.
curl -s https://agentleaderboards.com/api/v1/route/agentic.json | jq .picksNo key, open CORS, refreshed daily. Set your own budget and constraints →
Writing & editing code · Tool-call reliability · Following operator instructions · Recovering from shell errors · Hard reasoning · Mathematics · Precise instruction following · Long multi-turn chat · Long inputs · Creative writing · General assistant · Software & IT services · Legal & government · Medicine & healthcare · Science · Business & finance · Writing & language