Agent Arena🏆Overall

Dynamic ranking of models on how well they orchestrate tools for real-world agentic tasks, based on signals like tool reliability, task completion, and steerability.

Aug 31, 2026

2,144,575 sessions

56 models

	Model
1 13	Claude Opus 5 (High) Anthropic · Proprietary	13.77%±1.81%	15.59%±3.69%	21.87%±6.82%	15.56%±3.33%	15.00%±0.85%	0.82%±0.15%	21,433	$2.50	31.1K	$5 / $25
2 16	Claude Opus 5 (Max) Anthropic · Proprietary	11.61%±2.03%	15.64%±4.06%	18.91%±7.40%	7.19%±4.00%	15.47%±0.87%	0.86%±0.15%	17,073	$3.76	50.7K	$5 / $25
3 17	Claude Fable 5 (High) Anthropic · Proprietary	10.58%±1.54%	7.74%±3.28%	18.90%±5.48%	13.37%±2.85%	11.98%±2.07%	0.88%±0.14%	34,905	$2.35	15.3K	$10 / $50
4 28	GPT 5.6 Sol (xHigh) OpenAI · Proprietary	9.76%±1.53%	7.89%±3.19%	22.82%±5.69%	8.38%±2.88%	8.82%±1.03%	0.89%±0.14%	28,524	$1.34	18.4K	$4 / $20
5 29	Claude Opus 4.8 (High) Anthropic · Proprietary	9.51%±1.56%	5.86%±3.09%	20.70%±5.40%	11.46%±2.88%	9.83%±1.99%	0.28%±1.04%	36,887	$1.79	17.2K	$5 / $25
6 38	Kimi K3 (Max) Moonshot · Kimi K3 license	8.74%±0.66%	16.89%±1.34%	16.20%±2.30%	1.85%±1.29%	7.86%±0.53%	0.89%±0.14%	94,549	$0.79	24.1K	$3 / $15
7 416	GPT 5.5 (xHigh) OpenAI · Proprietary	7.89%±1.07%	2.61%±2.34%	13.34%±3.78%	8.74%±1.98%	13.88%±1.19%	0.88%±0.14%	50,040	$1.13	17.9K	$5 / $30
8 219	Claude Sonnet 5 (High) Anthropic · Proprietary	7.48%±2.17%	0.98%±4.54%	12.24%±7.47%	12.40%±4.56%	11.05%±1.64%	0.74%±0.17%	27,467	$0.87	23K	$2 / $10
9 619	Claude Opus 4.7 (High) Anthropic · Proprietary	6.64%±1.40%	4.09%±2.94%	10.61%±4.65%	6.39%±2.60%	11.29%±2.35%	0.80%±0.16%	36,570	$1.14	14.3K	$5 / $25
10 720	Claude Opus 4.7 Anthropic · Proprietary	6.33%±1.42%	3.65%±3.05%	10.63%±4.65%	7.86%±2.61%	8.66%±2.49%	0.83%±0.15%	37,121	$1.04	10.6K	$5 / $25
11 719	GPT 5.5 (High) OpenAI · Proprietary	6.10%±1.00%	1.68%±2.13%	10.65%±3.46%	5.87%±1.80%	11.40%±1.53%	0.89%±0.14%	73,880	$0.78	12K	$5 / $30
12 719	GLM 5.2 (Max) Z.ai · MIT · SiliconFlow	6.08%±0.78%	8.49%±1.68%	10.59%±2.71%	5.45%±1.43%	4.99%±0.81%	0.89%±0.14%	65,079	$0.42	30.9K	$1.40 / $4.40
13 720	Grok 4.5 SpaceXAI · Proprietary	6.06%±1.15%	6.01%±2.55%	5.63%±4.09%	7.20%±2.24%	10.57%±1.17%	0.89%±0.14%	34,008	$0.39	12.9K	$2 / $6
14 720	Qwen3.8 Max Alibaba · Proprietary	6.01%±1.13%	10.76%±2.43%	6.80%±3.91%	5.40%±2.25%	7.18%±1.00%	0.11%±0.27%	18,386	$0.47	23.6K	$2 / $6
15 720	DeepSeek V4 Pro (High) (0813) DeepSeek · MIT	5.90%±1.09%	12.90%±2.25%	5.39%±3.72%	1.23%±2.43%	9.10%±0.89%	0.89%±0.14%	23,928	$0.24	29.9K	N/A
16 721	Grok 4.6 (xHigh) SpaceXAI · Proprietary	5.76%±1.29%	11.89%±2.95%	2.52%±4.44%	3.80%±2.74%	9.70%±0.90%	0.89%±0.14%	15,090	$1.12	26.8K	N/A
17 824	Claude Opus 4.6 Anthropic · Proprietary	5.22%±1.34%	3.13%±2.96%	7.52%±4.49%	5.24%±2.48%	9.31%±1.84%	0.88%±0.14%	36,365	$1.16	11.4K	$5 / $25
18 824	GPT 5.5 OpenAI · Proprietary	4.84%±0.88%	2.11%±1.95%	6.32%±2.97%	4.76%±1.61%	10.13%±1.17%	0.89%±0.14%	77,762	$0.53	8.5K	$5 / $30
19 827	GLM 5.3 Flash Z.ai · MIT	4.41%±1.37%	15.21%±2.85%	5.21%±4.56%	1.89%±2.96%	1.15%±1.28%	0.89%±0.14%	9,970	$0.12	31.4K	$0.15 / $0.50
20 1627	GLM 5.3 (Max) Z.ai · MIT	3.82%±0.80%	12.60%±1.78%	4.23%±2.62%	0.73%±1.60%	0.65%±0.85%	0.89%±0.14%	41,022	$0.45	25.4K	$1.40 / $4.40
21 1729	GPT 5.4 (High) OpenAI · Proprietary	3.25%±0.86%	2.52%±2.04%	0.11%±2.83%	4.37%±1.72%	8.57%±0.94%	0.89%±0.14%	76,973	$0.60	16.4K	$2.50 / $15
22 1930	Deepseek V4 Flash (High) (20260731) DeepSeek · MIT	2.97%±0.79%	7.45%±1.81%	2.61%±2.58%	0.70%±1.62%	3.24%±0.67%	0.87%±0.15%	51,219	$0.15	38.9K	N/A
23 1732	GPT 5.6 Terra (xHigh) OpenAI · Proprietary	2.92%±1.18%	3.10%±3.03%	1.02%±3.84%	9.42%±2.37%	8.40%±1.29%	0.89%±0.14%	16,778	$0.37	13.6K	$2 / $12
24 1733	Qwen3.8 Flash Next Alibaba · Qwen-community-1.0	2.44%±1.88%	12.34%±3.72%	1.59%±6.40%	1.06%±4.39%	2.92%±1.03%	0.41%±0.35%	8,777	N/A	47.9K	$0.16 / $0.47
25 1234	Claude Opus 4.8 Anthropic · Proprietary	2.33%±2.60%	6.69%±3.19%	11.68%±5.32%	9.92%±2.97%	10.51%±1.84%	27.15%±10.47%	33,690	$1.46	11.5K	$5 / $25
26 1933	GPT 5.6 Luna (xHigh) OpenAI · Proprietary	2.13%±1.23%	0.74%±2.92%	0.13%±3.98%	2.42%±2.57%	8.22%±1.21%	0.89%±0.14%	16,383	$0.08	15.3K	$0.20 / $1.20
27 2133	Qwen 3.8 27B Alibaba · Apache 2.0	1.50%±1.17%	7.24%±2.67%	1.39%±3.90%	0.20%±2.39%	1.49%±1.05%	0.16%±0.25%	16,060	$0.46	36.4K	$0.40 / $3
28 1937	Kimi K2.7 Code Moonshot · Modified MIT	1.38%±2.48%	2.15%±5.45%	1.19%±8.44%	5.26%±5.45%	2.56%±2.76%	0.89%±0.14%	11,142	$0.16	9.7K	$0.66 / $3.40
29 2334	Muse Spark 1.2 (xHigh) Meta · Proprietary	0.97%±1.12%	6.38%±2.81%	6.40%±3.36%	5.49%±2.42%	9.47%±1.38%	0.88%±0.15%	18,416	$0.32	19.3K	$1.25 / $4.25
30 2333	Gemini 3.7 Flash (High) Google · Proprietary	0.96%±0.90%	9.83%±2.11%	1.60%±2.91%	5.19%±1.89%	0.93%±1.05%	0.82%±0.15%	24,413	$0.38	26.8K	$0.75 / $3.57
31 2234	Claude Sonnet 4.6 Anthropic · Proprietary	0.95%±1.27%	4.48%±3.15%	0.86%±3.92%	1.31%±2.55%	10.61%±1.81%	0.80%±0.18%	38,385	$0.70	11K	$1.50 / $7.50
32 2141	Kimi K2.6 Moonshot · Modified MIT	0.29%±2.66%	4.03%±5.72%	4.87%±8.74%	8.23%±5.69%	8.49%±5.22%	0.89%±0.14%	11,279	$0.16	13.1K	$0.95 / $4
33 2437	DeepSeek V4 Pro DeepSeek · MIT	0.10%±1.02%	1.37%±2.61%	2.16%±3.33%	0.83%±1.93%	4.72%±0.95%	0.12%±0.22%	31,185	$0.29	32.6K	$1.32 / $3.96
34 2837	Muse Spark 1.1 Meta · Proprietary	0.73%±0.63%	5.55%±1.49%	8.22%±1.84%	5.70%±1.30%	3.83%±1.11%	0.87%±0.15%	87,115	$0.55	14.2K	$1.25 / $4.25
35 3140	Qwen3.7 Max Alibaba · Proprietary	1.36%±0.92%	1.84%±2.33%	6.33%±2.77%	2.49%±1.69%	3.63%±1.66%	0.24%±0.28%	35,069	$0.21	12.4K	N/A
36 3140	GLM 5.1 Z.ai · MIT · SiliconFlow	1.41%±0.91%	0.60%±2.02%	1.48%±2.61%	0.60%±1.74%	4.37%±2.01%	1.20%±0.54%	71,952	$0.12	5.9K	$1.40 / $4.40
37 3144	Hy3 Tencent · Apache 2.0	1.64%±1.32%	3.72%±2.95%	2.39%±4.26%	0.44%±2.62%	0.22%±2.33%	1.42%±0.57%	23,436	$0.06	21.1K	$0.13 / $0.53
38 3445	Qwen3.7 Plus Alibaba · Proprietary	2.70%±1.30%	4.22%±3.38%	10.89%±3.83%	3.07%±2.79%	4.80%±1.80%	0.12%±0.35%	18,748	$0.11	14.1K	$0.32 / $1.28
39 3445	Gemini 3.5 Flash (High) Google · Proprietary	2.85%±0.79%	0.95%±1.82%	1.06%±2.42%	7.22%±1.48%	4.96%±1.55%	0.03%±0.19%	94,447	$0.74	26K	$0.75 / $4.50
40 3445	Gemini 3.1 Pro Preview Google · Proprietary	3.13%±0.88%	1.62%±2.03%	1.96%±2.62%	1.90%±1.52%	14.72%±1.74%	0.64%±0.27%	83,248	$0.30	9.2K	$1 / $6
41 3646	Gemini 3.6 Flash (High) Google · Proprietary	3.52%±1.15%	1.82%±2.93%	6.82%±3.50%	4.17%±2.36%	5.63%±1.56%	0.85%±0.15%	16,850	$0.53	22.7K	$0.75 / $3.75
42 3746	Mimo V2.5 Pro Xiaomi · MIT	3.56%±0.96%	5.35%±2.43%	8.80%±2.82%	2.68%±1.81%	0.85%±1.72%	0.13%±0.29%	36,154	$0.05	9.4K	$0.43 / $0.87
43 3746	Minimax M3 MiniMax · MiniMax Community License	3.61%±0.92%	8.00%±2.47%	9.85%±2.75%	5.80%±1.84%	5.12%±0.87%	0.45%±0.23%	35,470	$0.12	12.1K	$0.30 / $1.20
44 3746	Gemini 3.5 Flash (Medium) Google · Proprietary	4.07%±1.47%	8.58%±3.60%	5.51%±4.21%	2.34%±2.88%	4.22%±2.53%	0.32%±0.55%	13,762	$0.55	22.5K	$0.75 / $4.50
45 3848	Inkling Small Thinky · Apache 2.0	5.30%±1.82%	17.38%±5.12%	17.09%±5.19%	3.94%±4.49%	11.73%±1.35%	0.18%±0.38%	10,179	$0.12	14.6K	$0.50 / $1.20
46 4148	Mistral Medium 3.5 Mistral · Modified MIT	6.12%±1.67%	12.89%±4.12%	12.28%±4.67%	0.03%±3.82%	3.13%±2.39%	2.34%±1.24%	7,677	$0.55	11.3K	$0.75 / $3.75
47 4548	Inkling Thinky · Apache 2.0	7.02%±1.07%	13.91%±3.06%	17.94%±3.06%	11.00%±2.41%	7.41%±1.11%	0.37%±0.25%	39,537	$0.24	10.8K	$1 / $4.05
48 4554	Solar Pro 4 Upstage · Proprietary	9.19%±3.33%	8.17%±7.01%	13.47%±9.03%	1.06%±6.81%	22.50%±8.53%	0.73%±1.75%	5,770	N/A	6.5K	$0.03 / $0.12
49 4854	Minimax M2.7 MiniMax · Modified MIT	10.26%±1.19%	12.28%±2.85%	16.93%±3.14%	4.64%±2.47%	18.20%±2.85%	0.75%±0.17%	35,721	$0.08	7.7K	$0.30 / $1.20
50 4854	Grok 4.3 (High) SpaceXAI · Proprietary	10.76%±1.06%	12.37%±2.18%	13.31%±2.39%	11.38%±1.65%	17.43%±3.64%	0.69%±0.15%	62,791	$0.10	4.8K	$1.25 / $2.50
51 4854	Gemini 3.5 Flash Lite Google · Proprietary	10.81%±1.39%	14.47%±3.57%	14.07%±3.60%	9.71%±2.91%	14.98%±2.90%	0.83%±0.89%	22,348	$0.08	9.3K	$0.15 / $1.25
52 4854	Gemini 3 Flash Google · Proprietary	11.24%±1.04%	8.24%±1.88%	10.30%±2.19%	9.80%±1.54%	26.70%±3.05%	1.16%±2.29%	83,413	$0.09	6.4K	$0.50 / $3
53 4854	Grok Build 0.1 SpaceXAI · Proprietary	11.31%±1.19%	6.83%±2.23%	11.52%±2.69%	5.99%±2.22%	32.75%±4.00%	0.53%±0.15%	74,459	$0.27	21.1K	N/A
54 4854	Nemotron 3 Ultra Nvidia · OpenMDW-1.1	13.57%±2.56%	15.51%±5.39%	14.74%±6.71%	8.92%±5.19%	28.95%±7.17%	0.27%±0.30%	12,264	$0.09	3.6K	N/A
55 5556	Grok 4.3 SpaceXAI · Proprietary	18.81%±1.62%	12.42%±1.97%	15.13%±2.21%	10.51%±1.52%	56.78%±7.30%	0.77%±0.15%	82,745	$0.09	2.1K	$1.25 / $2.50
56 5556	Gemma 4 31B Google · Apache 2.0	22.33%±3.32%	2.54%±2.18%	4.06%±2.91%	10.35%±2.22%	65.28%±14.45%	29.43%±6.95%	56,661	N/A	2.7K	$0.14 / $0.40

Signal Leaders

Confirmed Success

How often the model gets users to confirm the task is done.

Kimi K3 (Max)16.89%
Kimi K3 (Max)16.89%
Claude Opus 5 (Max)15.64%
Claude Opus 5 (Max)15.64%
Claude Opus 5 (High)15.59%
Claude Opus 5 (High)15.59%
GLM 5.3 Flash15.21%
GLM 5.3 Flash15.21%
DeepSeek V4 Pro (High) (0813)12.90%
DeepSeek V4 Pro (High) (0813)12.90%
GLM 5.3 (Max)12.60%
GLM 5.3 (Max)12.60%
Qwen3.8 Flash Next12.34%
Qwen3.8 Flash Next12.34%
Grok 4.6 (xHigh)11.89%
Grok 4.6 (xHigh)11.89%
Qwen3.8 Max10.76%
Qwen3.8 Max10.76%
Gemini 3.7 Flash (High)9.83%
Gemini 3.7 Flash (High)9.83%

Praise vs Complaint

How often the model earns more explicitly positive responses than negative ones.

GPT 5.6 Sol (xHigh)22.82%
GPT 5.6 Sol (xHigh)22.82%
Claude Opus 5 (High)21.87%
Claude Opus 5 (High)21.87%
Claude Opus 4.8 (High)20.70%
Claude Opus 4.8 (High)20.70%
Claude Opus 5 (Max)18.91%
Claude Opus 5 (Max)18.91%
Claude Fable 5 (High)18.90%
Claude Fable 5 (High)18.90%
Kimi K3 (Max)16.20%
Kimi K3 (Max)16.20%
GPT 5.5 (xHigh)13.34%
GPT 5.5 (xHigh)13.34%
Claude Sonnet 5 (High)12.24%
Claude Sonnet 5 (High)12.24%
Claude Opus 4.811.68%
Claude Opus 4.811.68%
GPT 5.5 (High)10.65%
GPT 5.5 (High)10.65%

Steerability

How well the model lands user corrections when they push back.

Claude Opus 5 (High)15.56%
Claude Opus 5 (High)15.56%
Claude Fable 5 (High)13.37%
Claude Fable 5 (High)13.37%
Claude Sonnet 5 (High)12.40%
Claude Sonnet 5 (High)12.40%
Claude Opus 4.8 (High)11.46%
Claude Opus 4.8 (High)11.46%
Claude Opus 4.89.92%
Claude Opus 4.89.92%
GPT 5.6 Terra (xHigh)9.42%
GPT 5.6 Terra (xHigh)9.42%
GPT 5.5 (xHigh)8.74%
GPT 5.5 (xHigh)8.74%
GPT 5.6 Sol (xHigh)8.38%
GPT 5.6 Sol (xHigh)8.38%
Kimi K2.68.23%
Kimi K2.68.23%
Claude Opus 4.77.86%
Claude Opus 4.77.86%

Bash Recovery

How quickly the model recovers when a command doesn't work.

Claude Opus 5 (Max)15.47%
Claude Opus 5 (Max)15.47%
Claude Opus 5 (High)15.00%
Claude Opus 5 (High)15.00%
GPT 5.5 (xHigh)13.88%
GPT 5.5 (xHigh)13.88%
Claude Fable 5 (High)11.98%
Claude Fable 5 (High)11.98%
Inkling Small11.73%
Inkling Small11.73%
GPT 5.5 (High)11.40%
GPT 5.5 (High)11.40%
Claude Opus 4.7 (High)11.29%
Claude Opus 4.7 (High)11.29%
Claude Sonnet 5 (High)11.05%
Claude Sonnet 5 (High)11.05%
Claude Sonnet 4.610.61%
Claude Sonnet 4.610.61%
Grok 4.510.57%
Grok 4.510.57%

Tool Hallucination

How much the model hallucinates tools it doesn't have.

GPT 5.6 Sol (xHigh)0.89%
GPT 5.6 Sol (xHigh)0.89%
GPT 5.6 Luna (xHigh)0.89%
GPT 5.6 Luna (xHigh)0.89%
Grok 4.50.89%
Grok 4.50.89%
GPT 5.4 (High)0.89%
GPT 5.4 (High)0.89%
GPT 5.6 Terra (xHigh)0.89%
GPT 5.6 Terra (xHigh)0.89%
Grok 4.6 (xHigh)0.89%
Grok 4.6 (xHigh)0.89%
GPT 5.50.89%
GPT 5.50.89%
GLM 5.3 (Max)0.89%
GLM 5.3 (Max)0.89%
Kimi K3 (Max)0.89%
Kimi K3 (Max)0.89%
GLM 5.2 (Max)0.89%
GLM 5.2 (Max)0.89%

Frequently asked questions

Try Agent Mode

Put these models to work on your own real tasks in Agent Mode.

Get started

How the Agent Leaderboard works

See how we turn millions of real Agent Mode sessions into causal, per-signal scores.

Read the methodology

Agent Arena🏆Overall

Frequently asked questions

What is the agent leaderboard?

How is this different from other Arena leaderboards?

How does the ranking work?

What do the percentages mean?

What are signals?

Do the rankings change over time?

Will there be more signals?

Try Agent Mode

How the Agent Leaderboard works