Claude Code Opus 4.7 Performance Tracker | Marginlab

Track Claude Code's daily performance on SWE-Bench-Pro. Monitor for degradation with statistical significance testing.

Claude Code Opus 4.8 Performance Tracker | Marginlab Last updated: Jun 18, 2026 Claude Code Opus 4.8 Performance Tracker We are collecting a fresh Opus 4.8 baseline on SWE tasks before resuming statistical degradation detection. • Updated daily: Daily benchmarks on a curated subset of SWE-Bench-Pro • Collect baseline: New model baseline collection is active • What you see is what you get: We benchmark in Claude Code CLI with the SOTA model (currently Opus 4.8) directly, no custom harnesses. Receive Alerts About Powered by MARGIN EVALS New model - collecting baseline data. Degradation detection

saved by

• Emma Guo

related reading

• Best practices for Claude Code - Claude Code Docsanthropic.com
• Claude Statusstatus.claude.com
• Claude Code Cheat Sheetcc.storyfox.cz
• AI Model & API Providers Analysis | Artificial Analysisartificialanalysis.ai
• A Guide to Claude Code 2.0 and getting better at using coding agents – sankalp's blogsankalp.bearblog.dev
• Perplexityperplexity.ai
• GitHub - affaan-m/ECC: The agent harness performance optimization system. Skills, instincts, memory, security, and research-first development for Claude Code, Codex, Opencode, Cursor and beyond. · GitHubgithub.com
• GitHub - ccusage/ccusage: npx ccusage · GitHubgithub.com
• GitHub - brexhq/prompt-engineering: Tips and tricks for working with Large Language Models like OpenAI's GPT-4. · GitHubgithub.com
• Branches · HazyResearch/intelligence-per-watt · GitHubgithub.com
• Claude Code Hit Escape Velocity in January. So I Used Claude Code to /Catch-Up.zerotopete.com
• cchistory: Tracking Claude Code System Prompt and Tool Changesmariozechner.at