AI BriefingOpus 4.7 decline
·2026.04.17 00:12
Key point
In MRCR v2 8-needle at 1M tokens, Opus 4.7 scores 32.2%, significantly lower than 4.6's 78.3%.
Details
In MRCR v2 (8-needle) @ 1M tokens, Opus 4.7 (max)'s mean match ratio was 32.2%. Under the same conditions, Opus 4.6 (64k ext thinking) scored 78.3%.
Comparison values:
- GPT-5.4 (xhigh): 36.6%
- Gemini-3.1-Pro (thinking:high): 25.9%
The caption presents this result as a benchmark of Claude Opus 4.7's long context comprehension and sequential reasoning, showing that 4.7 significantly lags behind 4.6.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Step 5 Preview Released: 600B Parameter MoE Model Highlights Agentic Tasks and Financial Analysis Performance
StarCraft AI Benchmark 'Brood War Bench' Released... Only Codex Astra xhigh Achieves 18-0 RecordPrivatemode Benchmarks GLM-5.3-Flash as Vision-Capable Decision Model with Parity to Jev but Higher Cost