Agentic & Tool Use Leaderboard

Models ranked by Agentic Score, based on independent benchmark evaluations.

Each row shows the model's benchmark score alongside its pricing and output speed, so you can evaluate quality-to-cost tradeoffs at a glance.

Top 20 models ranked by agentic score

Rank	Model	Creator	Agentic Score	Price/1M	Speed
🥇	Anthropic: Claude Fable 5 Anthropic	Anthropic	68.2	$20.00	77 tok/s
🥈	Anthropic: Claude Opus 4.8 Anthropic	Anthropic	65.0	$10.00	58 tok/s
🥉	OpenAI: GPT-5.5 OpenAI	OpenAI	64.9	$11.25	82 tok/s
4	Anthropic: Claude Opus 4.7 Anthropic	Anthropic	63.6	$10.00	51 tok/s
5	GPT-5.5 (high) OpenAI	OpenAI	62.4	$11.25	83 tok/s
6	OpenAI: GPT-5.4 OpenAI	OpenAI	61.3	$5.63	166 tok/s
7	GPT-5.5 (medium) OpenAI	OpenAI	61.0	$11.25	71 tok/s
8	Google: Gemini 3.5 Flash Google	Google	60.2	$3.38	210 tok/s
9	Z.ai: GLM 5.2 Z AI	Z AI	60.0	$1.46	135 tok/s
10	Google: Gemini 3.1 Pro Preview Google	Google	57.7	$4.50	141 tok/s
11	Qwen: Qwen3.7 Max Alibaba	Alibaba	56.0	$1.88	196 tok/s
12	Claude Sonnet 4.6 (Adaptive Reasoning, Max Effort) Anthropic	Anthropic	55.1	$6.00	63 tok/s
13	GPT-5.5 (low) OpenAI	OpenAI	52.2	$11.25	70 tok/s
14	DeepSeek: DeepSeek V4 Pro DeepSeek	DeepSeek	51.9	$0.54	84 tok/s
15	MiniMax: MiniMax M3 MiniMax	MiniMax	51.5	$0.52	85 tok/s
16	MoonshotAI: Kimi K2.7 Code Kimi	Kimi	51.4	$1.43	52 tok/s
17	Xiaomi: MiMo-V2.5-Pro Xiaomi	Xiaomi	51.2	$0.54	46 tok/s
18	Muse Spark Meta	Meta	50.9	N/A	—
19	Nex AGI: Nex-N2-Pro Nex Agi	Nex Agi	50.1	$0.44	86 tok/s
20	MoonshotAI: Kimi K2.6 Kimi	Kimi	49.4	$1.35	54 tok/s