From a790ed94868a30ce58405bb93929e370c579c4b7 Mon Sep 17 00:00:00 2001
From: reacher-z <mtrxcop@gmail.com>
Date: Mon, 27 Jul 2026 19:46:11 +0800
Subject: [PATCH] docs: add ClawBench to related benchmarks

---
 README.md | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/README.md b/README.md
index acf2302..2c7fb12 100644
--- a/README.md
+++ b/README.md
@@ -33,6 +33,10 @@
 
 **Headline** — Best model-runtime pairing (Claude Opus 4.7 + Claude Code) = **41.2% PassRate**, vs >78% the same backbones reach on OSWorld-Verified.
 
+## Related Benchmarks
+
+- **[ClawBench](https://claw-bench.com/)** — A complementary live-web benchmark with 153 real-world tasks across 144 websites and safe request interception for reproducible evaluation. [Paper](https://arxiv.org/abs/2604.08523) · [Code](https://github.com/TIGER-AI-Lab/ClawBench)
+
 ## 🎬 Demo
 
 https://github.com/weavebench/WeaveBench/raw/main/docs/media/rabbitmq_dlq_demo.mp4