Scenario guide
Best AI models for Autonomous Agent / Tool Use
As of November 16, 2026, the highest-ranked model for Autonomous Agent / Tool Use is Gemini 4 Carbon, with a scenario score of 86.3 out of 100. The score blends this scenario's benchmark weights with its cost weight.
An agent that runs multi-step jobs on its own: shell commands, code changes, tool calls. Weighted toward agentic benchmarks that grade finished tasks (Terminal-Bench, SWE-bench) plus novel-problem reasoning, with a slice for self-correction because a long run fails if the model cannot recover from its own mistakes. Cost matters because agent loops resend large contexts on every step.
Rankings use the same scenario weights and cost blending as the interactive leaderboard on AI Model Analyzer. Data is min-max normalised per benchmark; missing scores are skipped without penalty.
- 1Gemini 4 CarbonGoogleScore 86.3Q 95.2In $1.00/M
- 2GPT-6.1 TerraOpenAIScore 84.1Q 90.0In $0.50/M
- 3GPT-6.1 SolOpenAIScore 83.0Q 91.1In $1.00/M
- 4GPT-5.3-CodexOpenAIScore 81.5Q 92.3In $1.75/M
- 5Gemini 3.1 ProGoogleScore 80.4Q 88.2In $1.00/M
- 6GPT-6 AstraOpenAIScore 79.9Q 93.1In $5.00/M
- 7GPT-5.5OpenAIScore 79.1Q 89.9In $2.50/M
- 8GPT-6 SolOpenAIScore 78.2Q 85.1In $1.00/M
- 9Claude Opus 5.5AnthropicScore 78.0Q 89.8In $4.00/M
- 10Gemini 4 ArgonGoogleScore 77.6Q 84.3In $1.00/M
- 11DeepSeek V4.1 Flash 1017DeepSeekScore 77.5Q 77.0In $0.15/M
- 12GPT-5.4OpenAIScore 77.2Q 85.0In $1.25/M
- 13Claude Sonnet 5.5AnthropicScore 77.2Q 86.3In $2.00/M
- 14Gemini 3.8 FlashGoogleScore 76.7Q 79.7In $0.38/M
- 15GPT-5.6 TerraOpenAIScore 76.7Q 83.6In $1.00/M