Best AI Models for Tool Use
Models with strong tool-use and function-calling support.
| # | Model | DFO Score ▾ | Tok/s | In $/1M | Out $/1M | Ctx | |
|---|---|---|---|---|---|---|---|
| 1 |  Claude Fable 5.1Anthropic | TOP PICK | 88.1 | 70.2 | $10.00 | $50.00 | 1M |
| 2 | Claude Opus 5Anthropic | 83.8 | 58.0 | $5.00 | $25.00 | 1M | |
| 3 |  Qwen3.8 MaxQwen | 74.9 | 37.6 | $2.00 | $6.00 | 1M | |
| 4 |  Muse Spark 1.3Meta | 79.3 | 139 | $1.25 | $4.25 | 1M | |
| 5 |  GLM 5.3Z.ai | 73.9 | 92.9 | $0.0790 | $1.00 | 1M | |
| 6 |  Grok 4.6SpaceXAI | 73.1 | 65.7 | $2.00 | $6.00 | 500K | |
| 7 |  GPT-6 AstraOpenAI | FRONTIER | 86.9 | 68.5 | $10.00 | $50.00 | 1.1M |
| 8 | Claude Fable 5Anthropic | 81.8 | 71.2 | $10.00 | $50.00 | 1M | |
| 9 | GPT-5.6 SolOpenAI | 77.5 | 89.2 | $4.00 | $20.00 | 1.1M | |
| 10 |  Kimi K3MoonshotAI | 71.9 | 43.1 | $0.2490 | $11.94 | 1M | |
| 11 | Claude Opus 5.5Anthropic | TOP PICKNEW | 95.0 | 97.8 | $4.00 | $20.00 | 1M |
| 12 | GPT-6 SolOpenAI | NEW | 78.5 | 121 | $2.00 | $10.00 | 1.1M |
| 13 | Grok 4.7SpaceXAI | NEW | 76.5 | 93.9 | $2.00 | $6.00 | 500K |
| 14 |  MiMo-V2.6-ProXiaomi | NEW | 76.4 | 50.4 | $0.4350 | $0.8700 | 1.1M |
| 15 |  Ling-2.6-1T (free)inclusionAI | BEST FOR AGENTS | 28.0 | 131 | Free | Free | 262K |
| 16 | Grok 4.20 Multi-Agent BetaxAI | BEST FOR AGENTS | 65.8 | 242 | $2.00 | $6.00 | 2M |
| 17 |  Hy3 preview (free)Tencent | 56.7 | 87.0 | Free | Free | 262K | |
| 18 | GLM 5.3 FlashZ.ai | BEST VALUE | 68.9 | 77.8 | $0.0480 | $0.2000 | 1M |
| 19 | Qwen3.8 2.4T A95BQwen | TOP PICK | 65.8 | 38.3 | $2.00 | $6.00 | 1M |
| 20 |  Nemotron 3 UltraNVIDIA | 51.3 | 195 | $0.6000 | $2.40 | 262K |
We deploy these models for businesses every week. Get a recommendation for your workload.
Get StartedModels with strong tool-use and function-calling support.
Leaderboards by use case
The overall table, re-ranked for the job you're hiring a model for.
How we rank AI models
The Design for Online AI Model Leaderboard ranks 843 models by the Artificial Analysis Intelligence Index, a composite of reasoning, knowledge, coding and agentic evaluations. The DFO score is each model's index as a share of the strongest model we track, so the leader scores 95 and every other model sits in proportion below it. No model scores 100. When Artificial Analysis stops scoring a model, its last index is scaled onto the current version so older and newer models are compared fairly.
Rankings come from the Artificial Analysis Intelligence Index; the Content Writing and SEO lists use Arena's blind human-preference ratings for writing and instruction following; pricing and availability come from the OpenRouter API. We deploy these models in client agents, chatbots and automations every week, and that experience shapes the written verdict on each model page.
Models are grouped into tiers (Frontier, Professional, Specialist, Efficient, Emerging and Legacy) to make like-for-like comparison easier, and newly released models are flagged so you can see what has just landed.
Leaderboard FAQ
How often is the leaderboard updated?
Pricing, availability and benchmark data are synced daily from our sources, and editorial scores are reviewed whenever a significant new model is released.
How is the overall score calculated?
The score is the model's Artificial Analysis Intelligence Index divided by the highest index on the board, times 95, so the current leader scores 95. Price, speed and our editorial notes are shown alongside it but do not change the order.
Where does the data come from?
Benchmarks from Artificial Analysis, writing and instruction-following ratings from Arena, pricing and availability from the OpenRouter API, and editorial notes from real deployments by the Design for Online team.