LLM Agent Benchmark Leaderboard

Name: LLM Agent Benchmark Leaderboard
Creator: DataLearner
License: https://creativecommons.org/licenses/by/4.0/

This page provides the LLM Agent benchmark leaderboard, covering Aider-Polyglot, τ²-Bench, Terminal Bench 2.0, Tool Decathlon, and OSWorld-Verified. Compare GPT, Claude, Qwen, and DeepSeek on tool use, task planning, and autonomous execution.

Updated on 2026-07-28 08:43:41

As of 2026-07, this page covers Aider-Polyglot, τ²-Bench, Terminal Bench 2.0, Tool Decathlon and related benchmarks for LLM Agent Benchmark Leaderboard, making it straightforward to compare within the same task family.

Click any model name to check context length, licensing, and pricing on its detail page. See Data Methodology for scoring details.

Benchmark

Agent能力评测Aider-Polyglot τ²-Bench

AI Agent - 工具使用Terminal Bench 2.0 Tool Decathlon OSWorld-Verified

More Benchmarks

Model Size:All 3B and below 7B 13B 34B 65B 100B and above

Model Type:All Reasoning Models Foundation Models Instruction/Chat Models Coding Models

Source:All Open Source Closed Source

Origin:All China

Model release cutoff:

Top picks

Ranked by Terminal Bench 2.0

Current SOTA

GPT-5.5

OpenAI

82.70Terminal Bench 2.0

View model

Best Open-Source

DeepSeek-V4-Pro

DeepSeek-AI

67.90Terminal Bench 2.0−14.80

View model

Best China-Made

Qwen3.7-Max-Preview

阿里巴巴

69.70Terminal Bench 2.0−13.00

View model

LLM Performance Results

Data source: DataLearnerAI

Click any row to open the model page. Tick the checkboxes to compare up to 4 models side by side.

Rank	Model						License
	GPT-5.5 Thinking EnabledTools OpenAI	—	—	82.70	—	78.70	Proprietary	Details
	Claude Mythos Preview Extended ThinkingTools Anthropic	—	—	82.00	—	79.60	Proprietary	Details
	GPT-5.3 Codex Standard ModeTools OpenAI	—	—	77.30	—	—	Proprietary	Details
4	GPT-5.4 Thinking Level · Extra HighTools OpenAI	—	—	75.10	—	75.00	Proprietary	Details
5	Qwen3.7-Max-Preview Thinking EnabledTools 阿里巴巴	—	—	69.70	—	—	Proprietary	Details
6	Opus 4.7 Extended ThinkingTools Anthropic	—	—	69.40	—	78.00	Proprietary	Details
7	Composer 2.5 Thinking Enabled Cursor	—	—	69.30	—	—	Proprietary	Details
8	Gemini 3.1 Pro Preview Thinking EnabledTools Google Deep Mind	—	90.80	68.50	—	—	Proprietary	Details
9	DeepSeek-V4-Pro Thinking Level · Extra HighTools DeepSeek-AI	—	—	67.90	—	—	Free commercial	Details
10	Kimi K2.6 Thinking EnabledTools Moonshot AI	—	—	66.70	50.00	73.10	Free commercial	Details
11	Qwen3.6-Max-Preview Deep Thinking ModeTools 阿里巴巴	—	—	65.40	—	—	Proprietary	Details
12	Claude Opus 4.6 Extended ThinkingTools Anthropic	—	91.89	65.40	—	72.70	Proprietary	Details
13	GLM 5.1 Thinking EnabledTools 智谱AI	—	—	63.50	40.70	—	Free commercial	Details
14	DeepSeek-V4-Pro Thinking EnabledTools DeepSeek-AI	—	—	63.30	—	—	Free commercial	Details
15	Composer 2 Thinking Enabled Cursor	—	—	61.70	—	—	Proprietary	Details
16	Qwen3.6-Max-Preview Thinking EnabledTools 阿里巴巴	—	—	61.60	—	—	Proprietary	Details
17	Qwen 3.6 Plus Preview Thinking EnabledTools 阿里巴巴	—	—	61.60	39.80	—	Proprietary	Details
18	GLM-5 Thinking EnabledTools 智谱AI	—	89.70	61.10	—	—	Free commercial	Details
19	GPT-5.4 mini Thinking Level · Extra HighTools OpenAI	—	—	60.00	42.90	72.10	Proprietary	Details
20	Qwen3.6-27B Thinking EnabledTools 阿里巴巴	—	—	59.30	—	—	Free commercial	Details
21	Opus 4.5 Extended ThinkingTools Anthropic	—	81.99	59.30	—	—	Proprietary	Details
22	DeepSeek-V4-Pro Standard ModeTools DeepSeek-AI	—	—	59.10	—	—	Free commercial	Details
23	Claude Sonnet 4.6 Thinking EnabledTools Anthropic	—	—	59.10	—	72.50	Proprietary	Details
24	Muse Spark Thinking EnabledTools Facebook AI研究实验室	—	—	59.00	—	—	Proprietary	Details
25	DeepSeek-V4-Flash Thinking Level · Extra HighTools DeepSeek-AI	—	—	56.90	—	—	Free commercial	Details
26	Gemini 3.0 Pro (Preview 11-2025) Thinking Level · HighTools Google Deep Mind	—	—	56.90	—	—	Proprietary	Details
27	DeepSeek-V4-Flash Thinking EnabledTools DeepSeek-AI	—	—	56.60	—	—	Free commercial	Details
28	Doubao Seed 2.0 Pro Thinking EnabledTools 字节跳动Seed团队	—	—	55.80	—	—	Proprietary	Details
29	Gemini 3.0 Pro (Preview 11-2025) Thinking EnabledTools Google Deep Mind	—	85.40	54.20	—	—	Proprietary	Details
30	Qwen3.5-397B-A17B Thinking EnabledTools 阿里巴巴	—	86.70	52.50	38.30	62.20	Free commercial	Details
31	MiniMax M2.5 Thinking EnabledTools MiniMaxAI	—	—	51.70	—	—	Free commercial	Details
32	Qwen3.6-35B-A3B Thinking Enabled 阿里巴巴	—	—	51.50	26.90	—	Free commercial	Details
33	Step 3.5 Flash Thinking EnabledTools StepFunAI	—	88.20	51.00	—	—	Free commercial	Details
34	Kimi K2.5 Thinking EnabledTools Moonshot AI	—	—	50.80	—	—	Free commercial	Details
35	DeepSeek-V4-Flash Standard ModeTools DeepSeek-AI	—	—	49.10	—	—	Free commercial	Details
36	Composer 1.5 Thinking Enabled Cursor	—	—	47.90	—	—	Proprietary	Details
37	M2.1 Thinking EnabledTools MiniMaxAI	—	—	47.90	—	—	Free commercial	Details
38	GPT-5.1 Thinking EnabledTools OpenAI	—	—	47.60	—	—	Proprietary	Details
39	Gemini 3.0 Flash Thinking EnabledTools Google Deep Mind	—	90.20	47.60	—	—	Proprietary	Details
40	DeepSeek V3.2 Thinking EnabledTools DeepSeek-AI	—	80.30	46.40	—	—	Free commercial	Details
41	GPT-5.4 nano Thinking Level · Extra HighTools OpenAI	—	—	46.30	35.50	39.00	Proprietary	Details
42	Claude Sonnet 4.5 Thinking EnabledTools Anthropic	—	84.70	42.80	—	61.40	Proprietary	Details
43	Qwen3.5-27B Thinking EnabledTools 阿里巴巴	—	79.00	41.60	—	56.20	Free commercial	Details
44	GLM-4.7 Thinking EnabledTools 智谱AI	—	87.40	41.00	—	—	Free commercial	Details
45	Composer 1 Thinking Enabled Cursor	—	—	40.00	—	—	Proprietary	Details
46	Qwen3-Coder-Next Standard ModeTools 阿里巴巴	—	—	36.20	—	—	Free commercial	Details
47	Gemini 2.5-Pro Thinking EnabledTools Google Deep Mind	—	—	32.60	—	—	Proprietary	Details
48	GPT-5 Thinking Enabled OpenAI	88.00	—	—	—	—	Proprietary	Details
49	GPT-5 Thinking Enabled OpenAI	86.70	—	—	—	—	Proprietary	Details
50	o3-pro Thinking Enabled OpenAI	84.90	—	—	—	—	Proprietary	Details

GPT-5.5 OpenAI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.082.70

Tool Decathlon—

OSWorld-Verified78.70

Proprietary

Claude Mythos Preview Anthropic

Extended ThinkingTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.082.00

Tool Decathlon—

OSWorld-Verified79.60

Proprietary

GPT-5.3 Codex OpenAI

Standard ModeTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.077.30

Tool Decathlon—

OSWorld-Verified—

Proprietary

GPT-5.4 OpenAI

Thinking Level · Extra HighTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.075.10

Tool Decathlon—

OSWorld-Verified75.00

Proprietary

Qwen3.7-Max-Preview 阿里巴巴

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.069.70

Tool Decathlon—

OSWorld-Verified—

Proprietary

Opus 4.7 Anthropic

Extended ThinkingTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.069.40

Tool Decathlon—

OSWorld-Verified78.00

Proprietary

Composer 2.5 Cursor

Thinking Enabled

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.069.30

Tool Decathlon—

OSWorld-Verified—

Proprietary

Gemini 3.1 Pro Preview Google Deep Mind

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench90.80

Terminal Bench 2.068.50

Tool Decathlon—

OSWorld-Verified—

Proprietary

DeepSeek-V4-Pro DeepSeek-AI

Thinking Level · Extra HighTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.067.90

Tool Decathlon—

OSWorld-Verified—

Free commercial

Kimi K2.6 Moonshot AI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.066.70

Tool Decathlon50.00

OSWorld-Verified73.10

Free commercial

Qwen3.6-Max-Preview 阿里巴巴

Deep Thinking ModeTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.065.40

Tool Decathlon—

OSWorld-Verified—

Proprietary

Claude Opus 4.6 Anthropic

Extended ThinkingTools

Aider-Polyglot—

τ²-Bench91.89

Terminal Bench 2.065.40

Tool Decathlon—

OSWorld-Verified72.70

Proprietary

GLM 5.1 智谱AI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.063.50

Tool Decathlon40.70

OSWorld-Verified—

Free commercial

DeepSeek-V4-Pro DeepSeek-AI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.063.30

Tool Decathlon—

OSWorld-Verified—

Free commercial

Composer 2 Cursor

Thinking Enabled

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.061.70

Tool Decathlon—

OSWorld-Verified—

Proprietary

Qwen3.6-Max-Preview 阿里巴巴

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.061.60

Tool Decathlon—

OSWorld-Verified—

Proprietary

Qwen 3.6 Plus Preview 阿里巴巴

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.061.60

Tool Decathlon39.80

OSWorld-Verified—

Proprietary

GLM-5 智谱AI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench89.70

Terminal Bench 2.061.10

Tool Decathlon—

OSWorld-Verified—

Free commercial

GPT-5.4 mini OpenAI

Thinking Level · Extra HighTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.060.00

Tool Decathlon42.90

OSWorld-Verified72.10

Proprietary

Qwen3.6-27B 阿里巴巴

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.059.30

Tool Decathlon—

OSWorld-Verified—

Free commercial

Opus 4.5 Anthropic

Extended ThinkingTools

Aider-Polyglot—

τ²-Bench81.99

Terminal Bench 2.059.30

Tool Decathlon—

OSWorld-Verified—

Proprietary

DeepSeek-V4-Pro DeepSeek-AI

Standard ModeTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.059.10

Tool Decathlon—

OSWorld-Verified—

Free commercial

Claude Sonnet 4.6 Anthropic

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.059.10

Tool Decathlon—

OSWorld-Verified72.50

Proprietary

Muse Spark Facebook AI研究实验室

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.059.00

Tool Decathlon—

OSWorld-Verified—

Proprietary

DeepSeek-V4-Flash DeepSeek-AI

Thinking Level · Extra HighTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.056.90

Tool Decathlon—

OSWorld-Verified—

Free commercial

Gemini 3.0 Pro (Preview 11-2025)Google Deep Mind

Thinking Level · HighTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.056.90

Tool Decathlon—

OSWorld-Verified—

Proprietary

DeepSeek-V4-Flash DeepSeek-AI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.056.60

Tool Decathlon—

OSWorld-Verified—

Free commercial

Doubao Seed 2.0 Pro 字节跳动Seed团队

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.055.80

Tool Decathlon—

OSWorld-Verified—

Proprietary

Gemini 3.0 Pro (Preview 11-2025)Google Deep Mind

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench85.40

Terminal Bench 2.054.20

Tool Decathlon—

OSWorld-Verified—

Proprietary

Qwen3.5-397B-A17B 阿里巴巴

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench86.70

Terminal Bench 2.052.50

Tool Decathlon38.30

OSWorld-Verified62.20

Free commercial

MiniMax M2.5 MiniMaxAI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.051.70

Tool Decathlon—

OSWorld-Verified—

Free commercial

Qwen3.6-35B-A3B 阿里巴巴

Thinking Enabled

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.051.50

Tool Decathlon26.90

OSWorld-Verified—

Free commercial

Step 3.5 Flash StepFunAI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench88.20

Terminal Bench 2.051.00

Tool Decathlon—

OSWorld-Verified—

Free commercial

Kimi K2.5 Moonshot AI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.050.80

Tool Decathlon—

OSWorld-Verified—

Free commercial

DeepSeek-V4-Flash DeepSeek-AI

Standard ModeTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.049.10

Tool Decathlon—

OSWorld-Verified—

Free commercial

Composer 1.5 Cursor

Thinking Enabled

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.047.90

Tool Decathlon—

OSWorld-Verified—

Proprietary

M2.1 MiniMaxAI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.047.90

Tool Decathlon—

OSWorld-Verified—

Free commercial

GPT-5.1 OpenAI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.047.60

Tool Decathlon—

OSWorld-Verified—

Proprietary

Gemini 3.0 Flash Google Deep Mind

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench90.20

Terminal Bench 2.047.60

Tool Decathlon—

OSWorld-Verified—

Proprietary

DeepSeek V3.2 DeepSeek-AI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench80.30

Terminal Bench 2.046.40

Tool Decathlon—

OSWorld-Verified—

Free commercial

GPT-5.4 nano OpenAI

Thinking Level · Extra HighTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.046.30

Tool Decathlon35.50

OSWorld-Verified39.00

Proprietary

Claude Sonnet 4.5 Anthropic

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench84.70

Terminal Bench 2.042.80

Tool Decathlon—

OSWorld-Verified61.40

Proprietary

Qwen3.5-27B 阿里巴巴

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench79.00

Terminal Bench 2.041.60

Tool Decathlon—

OSWorld-Verified56.20

Free commercial

GLM-4.7 智谱AI

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench87.40

Terminal Bench 2.041.00

Tool Decathlon—

OSWorld-Verified—

Free commercial

Composer 1 Cursor

Thinking Enabled

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.040.00

Tool Decathlon—

OSWorld-Verified—

Proprietary

Qwen3-Coder-Next 阿里巴巴

Standard ModeTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.036.20

Tool Decathlon—

OSWorld-Verified—

Free commercial

Gemini 2.5-Pro Google Deep Mind

Thinking EnabledTools

Aider-Polyglot—

τ²-Bench—

Terminal Bench 2.032.60

Tool Decathlon—

OSWorld-Verified—

Proprietary

GPT-5 OpenAI

Thinking Enabled

Aider-Polyglot88.00

τ²-Bench—

Terminal Bench 2.0—

Tool Decathlon—

OSWorld-Verified—

Proprietary

GPT-5 OpenAI

Thinking Enabled

Aider-Polyglot86.70

τ²-Bench—

Terminal Bench 2.0—

Tool Decathlon—

OSWorld-Verified—

Proprietary

o3-pro OpenAI

Thinking Enabled

Aider-Polyglot84.90

τ²-Bench—

Terminal Bench 2.0—

Tool Decathlon—

OSWorld-Verified—

Proprietary

Sort by:

Showing 50 of 149 modelsView Terminal Bench 2.0 benchmark page