Best model for your job
Best AI model for agents
For AI agents, Claude Opus 5 scores highest on the current data (64.6), weighting agentic & tool use 60%, instruction following 20%, reasoning 20%. The cheapest model in the top 10 is Gemini 3.8 Flash at $0.75 / $3.75 per million tokens.
Last verified
Weights: Agentic & Tool Use 60%, Instruction Following 20%, Reasoning 20%. Agents must plan, call tools and recover from errors over many steps.
| # | Model | Provider | Fit | Agentic & Tool Use | Instruction Following | Reasoning | Input $/M | Output $/M |
|---|---|---|---|---|---|---|---|---|
| 1 | Claude Opus 5 | Anthropic | 64.6 | 55.6 | 79.2 | 77.2 | $5 | $25 |
| 2 | GPT-6 Astra | OpenAI | 64.0 | 52.9 | 76.3 | 85.1 | $10 | $50 |
| 3 | Claude Fable 5 | Anthropic | 63.5 | 54.0 | 78.6 | 76.8 | $10 | $50 |
| 4 | Claude Fable 5.1 | Anthropic | 61.6 | 50.7 | 79.2 | 76.7 | $10 | $50 |
| 5 | GPT-5.6 Sol | OpenAI | 60.7 | 50.3 | 77.7 | 74.8 | $4 | $20 |
| 6 | GPT-5.5 | OpenAI | 60.5 | 50.7 | 77.5 | 72.8 | $5 | $30 |
| 7 | Claude Opus 5.5 | Anthropic | 59.2 | 45.3 | 80.0 | 80.2 | $4 | $20 |
| 8 | Claude Opus 4.6 | Anthropic | 58.1 | 51.1 | 79.5 | 57.8 | $5 | $25 |
| 9 | Claude Opus 4.8 | Anthropic | 57.0 | 47.6 | 77.4 | 64.7 | $5 | $25 |
| 10 | Gemini 3.8 Flash | 56.1 | 41.8 | 78.0 | 76.9 | $0.75 | $3.75 | |
| 11 | GPT-5.4 | OpenAI | 55.7 | 46.5 | 77.1 | 61.8 | $2.50 | $15 |
| 12 | GPT-6.1 Sol | OpenAI | 55.5 | 39.6 | 77.0 | 81.9 | $2 | $10 |
| 13 | Claude Opus 4.7 | Anthropic | 55.2 | 47.9 | 78.4 | 53.8 | $5 | $25 |
| 14 | Gemini 3.7 Flash | 54.8 | 42.1 | 77.7 | 70.0 | $0.75 | $3.75 | |
| 15 | Gemini 4 Argon | 54.4 | 49.2 | 80.1 | 44.2 | — | — | |
| 16 | Qwen3.8 Max | Alibaba (Qwen) | 53.6 | 45.4 | 77.6 | 54.4 | $2 | $6 |
| 17 | Claude Sonnet 5.5 | Anthropic | 53.5 | 45.0 | 78.3 | 54.0 | $2 | $10 |
| 18 | Kimi K3 (open weights) | Moonshot AI | 53.3 | 41.8 | 77.7 | 63.0 | $3 | $15 |
| 19 | Grok 4.5 | xAI | 53.0 | 44.4 | 76.0 | 56.1 | $2 | $6 |
| 20 | Claude Opus 4.5 | Anthropic | 52.4 | 47.3 | 77.5 | 42.6 | $5 | $25 |
| 21 | Gemini 3.1 Pro Preview | 52.4 | 37.7 | 77.0 | 71.7 | $2 | $12 | |
| 22 | GPT-6 Sol | OpenAI | 52.0 | 37.2 | 74.5 | 74.0 | $2 | $10 |
| 23 | GPT-5.6 Terra | OpenAI | 51.5 | 40.1 | 76.4 | 60.7 | $2 | $12 |
| 24 | Grok 4.6 | xAI | 51.0 | 39.4 | 75.4 | 61.4 | $2 | $6 |
| 25 | Claude Sonnet 5 | Anthropic | 50.7 | 42.8 | 76.3 | 49.1 | $2 | $10 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Top two head to head: Claude Opus 5 vs GPT-6 Astra
Frequently asked questions
What is the best ai model for agents?
For AI agents, Claude Opus 5 scores highest on the current data (64.6), weighting agentic & tool use 60%, instruction following 20%, reasoning 20%. The cheapest model in the top 10 is Gemini 3.8 Flash at $0.75 / $3.75 per million tokens.
How is this shortlist built?
Agents must plan, call tools and recover from errors over many steps. Each model's category scores are blended with those weights; only ranked models with results in every needed category are listed.
Other jobs
- Best AI model for coding
- Best AI model for debugging
- Best AI model for building websites
- Best local LLM for coding
- Best AI model for RAG
- Best AI model for research
- Best AI model for studying
- Best AI model for math
- Best AI model for hard reasoning
- Best AI model for data analysis
- Best AI model for Excel
- Best AI model for data extraction
- Best AI model for writing
- Best AI model for emails
- Best AI model for copywriting
- Best AI model for creative writing
- Best AI model for translation
- Best AI model for language learning
- Best AI model for customer service
- Best AI model for presentations
- Best AI model for PDFs
- Best AI assistant for everyday use