Skip to content

Benchmark

Apex Agents

Agentic benchmarks measure how well models carry out tasks that take several steps. Scores are percentages. Higher is better.

Models measured
72
Scores recorded
83
Sources
3
IQ weight
2.12

Score distribution

50.7%
Mean
51.6%
Median
16.9%
Std dev
27.9%
10th percentile
70%
90th percentile
4.4%
Min
4.4%84.2%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.12 for chat models, measured over 72 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Inkling 2 Small· Epoch AI Benchmarking Hub data46.8%
52Claude Opus 4.6max· Epoch AI Benchmarking Hub data46.3%
53Muse Spark 1.2· Epoch AI Benchmarking Hub data45.4%
54GLM 5.2· Epoch AI Benchmarking Hub data45.2%
55Claude Sonnet 5.5medium· Epoch AI Benchmarking Hub data44.6%
56GPT-6 Lunamax· Epoch AI Benchmarking Hub data44.3%
57Claude Sonnet 4.6high· Epoch AI Benchmarking Hub data43%
58GPT-5.6 Luna· Epoch AI Benchmarking Hub data43%
59Kimi K3max effort · temp 1.0 · top_p 1.0· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.41%
60GLM 5.1· Epoch AI Benchmarking Hub data40.9%
61DeepSeek V4.1 Flash· Epoch AI Benchmarking Hub data39.5%
62MiniMax M3· Epoch AI Benchmarking Hub data
Values from 2 sources
37.7%
63Kimi K2.7 Code· Epoch AI Benchmarking Hub data37.6%
64Hy4 preview· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim37.1%
65Hy4 previewHF Eval Result· Reported by Tencent Hy· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.37.1%
66Gemini 3.1 Pro Preview· Epoch AI Benchmarking Hub data
Values from 2 sources
35.1%
67Doubao Seed 2.1 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim33.8%
68Inkling FP4· Epoch AI Benchmarking Hub data33.8%
69Gemini 3.1 Pro Previewsetting not named by the source· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.33.5%
70Muse Spark Open 816B A42B· Epoch AI Benchmarking Hub data33.4%
71Muse Spark 1.1· Epoch AI Benchmarking Hub data31.8%
72Gemini 3.5 Flash Lite· Epoch AI Benchmarking Hub data29.3%
73Doubao Seed 2.1 Turbo· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim29.2%
74Ling 3.0 Flash Fin· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim29.2%
75Kimi K2.6· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim27.9%
76Kimi K2.6Kimi K2.6 official benchmark table · thinking · temp 1.0 · top_p 1.0 · 262,144 context · 452-task subset· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.27.9%
77MiniMax M3Archipelago · ReAct Toolbelt · Claude Sonnet 4.6 grader· Reported by MiniMax Research· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.27.7%
78Gemini 3.5 Flash· Epoch AI Benchmarking Hub data27.5%
79Hy3 (Deep Infra)· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim25.6%
80Qwen3.5 397B A17B· Epoch AI Benchmarking Hub data24.9%
81Nemotron 3 Ultra· Epoch AI Benchmarking Hub data22.7%
82DeepSeek V3.2· Epoch AI Benchmarking Hub data21.3%
83GPT OSS 120B· Epoch AI Benchmarking Hub data4.4%

Showing 51–83 of 83 results.

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · Data by LLM Stats · BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.