Benchmark
WebArena
Agentic benchmarks measure how well models carry out tasks that take several steps. Scores are percentages. Higher is better.
- Models measured
- 9
- Scores recorded
- 12
- Sources
- 2
- IQ weight
- 1.20
Score distribution
- 45.4%
- Mean
- 58.9%
- Median
- 27.3%
- Std dev
- 7.1%
- 10th percentile
- 71.2%
- 90th percentile
- 1.4%
- Min
This benchmark has a weight of 1.20 for chat models, measured over 9 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.5pass@4 · Computer Use API + browser tools · single policy · general prompts · official grader· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 72.4% |
| 2 | Claude Opus 4.5pass@3 · Computer Use API + browser tools · single policy · general prompts · official grader· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 71.2% |
| 3 | Claude Opus 4.5pass@2 · Computer Use API + browser tools · single policy · general prompts · official grader· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 69.5% |
| 4 | Claude Sonnet 4.6single policy · screenshot + DOM navigation · general prompts · Average@5· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 65.6% |
| 5 | Claude Opus 4.5pass@1 · Computer Use API + browser tools · single policy · general prompts · official grader · avg@5· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 65.3% |
| 6 | Kimi K2.5· Steel.dev Agent Leaderboards data | 58.9% |
| 7 | Claude Sonnet 4.5pass@1 · Computer Use API + browser tools · single policy · general prompts · official grader · avg@5· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 58.5% |
| 8 | Claude Haiku 4.5pass@1 · Computer Use API + browser tools · single policy · general prompts · official grader · avg@5· Reported by Anthropic· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it. | 53.1% |
| 9 | GPT 4o· Steel.dev Agent Leaderboards data | 13.1% |
| 10 | GPT 35 Turbo 16k· Steel.dev Agent Leaderboards data | 8.9% |
| 11 | Gemini Pro· Steel.dev Agent Leaderboards data | 7.1% |
| 12 | mixtral-8x7b· Steel.dev Agent Leaderboards data | 1.4% |
Data credits: 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. · Steel.dev Agent Leaderboards, MIT (licence)