ClawBench
ClawBench is an open benchmark for browser AI agents on everyday tasks, measuring how well agents browse and complete real workflows.
Updated
✨ Try AgentScout's free AI image generator →Good for: browsing the web, automating workflows, evaluating agents
Use cases
About ClawBench
ClawBench is listed in our evals & observability category and is aimed at builders. ClawBench is an open benchmark for browser AI agents on everyday tasks, measuring how well agents browse and complete real workflows. Pricing, as the maker states it: Open source. Maturity: Active. We last checked the official link on 2026-11-07.
ClawBench at a glance
- Category: Evals & observability
- Built for: Developers and teams building with AI agents
- Pricing: Open source (as the maker states it)
- GitHub: TIGER-AI-Lab/ClawBench · 1,223 stars (as of Nov 7, 2026)
- Official site: claw-bench.com
- Listed on AgentScout: Nov 7, 2026
ClawBench FAQ
What is ClawBench?
ClawBench is an open benchmark for browser AI agents on everyday tasks, measuring how well agents browse and complete real workflows.
How much does ClawBench cost?
ClawBench's pricing, as the maker states it: Open source. Open-source tools are free to use, though you may still pay for model API usage or a hosted plan. Check the official site for current plans.
Where can I find ClawBench's source code?
On GitHub at TIGER-AI-Lab/ClawBench, which had 1,223 stars as of Nov 7, 2026.
What are alternatives to ClawBench?
Similar tools include Galileo Agent Leaderboard, OSWorld, InferenceX, Userpilot Agent Analytics, Kore.ai Autoloop and Proton.
Facts from our link-checked listing, the GitHub repository and the official site. See something wrong? Suggest an edit.
ClawBench ratings and reviews
Be the first to rate ClawBench: tap 1 to 5 stars below. A written review is optional.
ClawBench alternatives
Other evals & observability tools in our directory. See all evals & observability tools →
Galileo Agent Leaderboard
Galileo's leaderboard that ranks how well LLMs perform as agents on tool use and multi-step tasks.
OSWorld
Benchmark for multimodal computer-use agents on open-ended tasks in real desktop environments.
InferenceX
InferenceX is SemiAnalysis' open-source inference research platform that continuously benchmarks how fast and how efficiently AI models are served on different accelerators and software stacks. It runs end-to-end serving benchmarks for new models from day zero on NVIDIA B200, B300 and GB200 and AMD MI355X systems with vLLM and SGLang, adds networking, kernel and hardware microbenchmarks, measures power draw, and publishes results on inferencex.com. Apache 2.0.
Userpilot Agent Analytics
Userpilot Agent Analytics shows product teams how users actually get on with the AI agent inside their app. It reads the agent's conversations at scale and reports whether people got what they needed or left frustrated, so teams can see where the agent fails, which requests it can't handle and how agent use ties back to product engagement, alongside Userpilot's existing product analytics.
Kore.ai Autoloop
Optimization engine in Kore.ai Agent Platform Artemis that evaluates agents against business goals, diagnoses failures from traces, applies targeted fixes, and re-verifies; GA Oct 2026.
Proton
The Fastest Unified Streaming SQL Engine in a Single C++ Binary. Millisecond latency.
Get your badge
Built ClawBench? Optional. Showing this badge does not change your rank or earn a Featured spot. It links to this page. About badges