agents benchmark
SeedClawBench
SeedClawBench is an agentic coding benchmark measuring overall model performance on real-world, tool-using software development tasks.
Updated Aug 7, 2026
Published models2
Registry coverage2
MetricScore
EvidenceB
SeedClawBench leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | BY | 66.6% | 100.0% | 2 | C | |
| 2 | BY | 63.8% | 0.0% | 2 | C |
Score distribution
Top published rows on the benchmark's original scale.
SeedClawBench highlights
The top published results on this benchmark's own scale.
What is SeedClawBench?
Definition and scoring fields from the benchmark registry.
SeedClawBench is an agentic coding benchmark measuring overall model performance on real-world, tool-using software development tasks.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- SeedClawBench
- Modality
- text
- Primary category
- agents
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- seedclawbench|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about SeedClawBench.
Which model scores highest on SeedClawBench?
Seed 2.1 Pro is currently ranked first with 66.6%.
What does SeedClawBench measure?
SeedClawBench is an agentic coding benchmark measuring overall model performance on real-world, tool-using software development tasks.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
2 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.