Anthropic
Claude Opus 5.5
Part of Anthropic’s Claude 5.5 family. Registry facts are limited to what Anthropic states on its own pages, each linked to where we found it. Benchmark results carry their own source and provenance label.
CurrentVerified against official sources
Verified facts
“source” links to the official page each value was confirmed on
- Organization
- Anthropic
- Family
- Claude 5.5
- Release date
- (released)source for Release date (external site)
- Context window
- 1,000,000 tokensStated as “1M tokens”source for Context window (external site)
- Max output tokens
- 128,000 tokensStated as “128K tokens (synchronous Messages API; up to 300K on Message Batches API with beta header output-300k-2026-03-24)”source for Max output tokens (external site)
- Knowledge cutoff
- 2026-06 (reliable knowledge cutoff)source for Knowledge cutoff (external site)
- Input modalities
- Text, Imagesource for Input modalities (external site)
- Output modalities
- Textsource for Output modalities (external site)
- Weights
- API-servedsource for Weights (external site)
- License
- Not verified
- API model ID
claude-opus-5-5source for API model ID (external site)- Status
- Current
- Last verified
Reported benchmark results
2 results on file, all reported by Anthropic. Scores are copied exactly as published, with the settings the source stated. Rows are different measurements — compare a score only with others in the same benchmark variant, via the variant link.
| Benchmark | Variant | Model detail | Score | Settings (as stated) | Source | Evidence |
|---|---|---|---|---|---|---|
| HLE | With tools — see all HLE results for this variant | adaptive thinking at max effort | 67.7%No CI reported | with tools; adaptive thinking at max effort (page default for all Opus 5.5 results unless otherwise noted). Anthropic evaluated Opus 5.5 with production safeguards enabled; when they intervened, biology and frontier LLM development tasks were completed by Claude Opus 5 and cybersecurity tasks by Claude Opus 4.8 (page note). | Show excerpt for Claude Opus 5.5 (adaptive thinking at max effort) on HLE, With tools“Humanity's Last Exam 67.7% with tools” Retrieved | |
| Terminal-Bench | 4.0 — see all Terminal-Bench results for this variant | xhigh effort | 66.4%No CI reported | xhigh effort (per page footnote); stated standard error ±2.6 pts. Anthropic evaluated Opus 5.5 with production safeguards enabled; when they intervened, biology and frontier LLM development tasks were completed by Claude Opus 5 and cybersecurity tasks by Claude Opus 4.8 (page note). | Vendor-reportedAnthropicanthropic.com — source for Claude Opus 5.5 (xhigh effort) on Terminal-Bench, 4.0 (external site) | Show excerpt for Claude Opus 5.5 (xhigh effort) on Terminal-Bench, 4.0“Agentic coding Terminal-Bench 4.0¹ 66.4%” Retrieved |
- Vendor-reported:
- Published by the organisation that built the model. Not independently reproduced.