agents benchmark
WildClawBench
WildClawBench is an agentic coding benchmark from InternLM/Claw-Eval that reports overall model performance on real-world tool-using development tasks.
Updated Aug 7, 2026
Published models4
Registry coverage4
MetricScore
EvidenceB
WildClawBench leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | BY | 62.8% | 100.0% | 4 | C | |
| 2 | BY | 61.7% | 66.7% | 4 | C | |
| 3 | TE | 53.6% | 33.3% | 4 | C | |
| 4 | XI | 43.0% | 0.0% | 4 | C |
Score distribution
Top published rows on the benchmark's original scale.
WildClawBench highlights
The top published results on this benchmark's own scale.
What is WildClawBench?
Definition and scoring fields from the benchmark registry.
WildClawBench is an agentic coding benchmark from InternLM/Claw-Eval that reports overall model performance on real-world tool-using development tasks.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- WildClawBench
- Modality
- text
- Primary category
- agents
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- wildclawbench|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about WildClawBench.
Which model scores highest on WildClawBench?
Seed 2.1 Turbo is currently ranked first with 62.8%.
What does WildClawBench measure?
WildClawBench is an agentic coding benchmark from InternLM/Claw-Eval that reports overall model performance on real-world tool-using development tasks.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
4 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.