Claude Opus 4.7
24 values about Claude Opus 4.7 from 3 documents, in 6 metric families. In each family, values from its own document come first.
- Developer
- Anthropic
- Release date
- 16 Apr 2026The date of its first card in the dataset, the Claude Opus 4.7 System Card.
- Availability
- Public
- Its own document
- Claude Opus 4.7 System Card · 16 Apr 2026
- Also reported in
- Claude Opus 4.8 System Card · 28 May 2026
- Claude Opus 5 System Card · 24 Jul 2026
Data as of 26 Sep 2026 · Dataset v0.1 · Methodology v0.1 · Changelog
Values by metric family
Each value as the document printed it. Select a value for its source, its checks and its history. Values from other documents are listed after the model’s own, each marked as the first report of that measure or as a restatement.
KeyVerifiedUnverifiedDisputedCorrected read off a figure or stated in wordsA value opens its source and history.
M1 · Evaluation awareness 1 value
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentClaude Opus 4.7 System Card · 16 Apr 2026 | ||
| Automated behavioral auditShare of transcripts verbalizing eval awareness | Claude Opus 4.7 System Card16 Apr 2026 | |
M2 · Reward hacking 1 value
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentClaude Opus 4.7 System Card · 16 Apr 2026 | ||
| Impossible tasksReward hack rate | Claude Opus 4.7 System Card16 Apr 2026 | |
M3 · Sabotage and sandbagging 2 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentClaude Opus 4.7 System Card · 16 Apr 2026 | ||
| Minimal-LinuxBenchStealth success rateCondition: CoT visible to monitor | Claude Opus 4.7 System Card16 Apr 2026 | |
| SHADE-ArenaStealth success rateCondition: extended thinking | Claude Opus 4.7 System Card16 Apr 2026 | |
M7 · Harmful compliance and over-refusal 3 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentClaude Opus 4.7 System Card · 16 Apr 2026 | ||
| Single-turn benign requestsOver-refusal rate | Claude Opus 4.7 System Card16 Apr 2026 | |
| From other documents | ||
| Single-turn benign requestsOver-refusal rateCondition: no system prompt; APIClaude Opus 4.8 System Card28 May 2026First reported | Claude Opus 4.8 System Card28 May 2026First reported | |
| Single-turn violative requestsHarmless response rateCondition: no system prompt; APIClaude Opus 4.8 System Card28 May 2026First reported | Claude Opus 4.8 System Card28 May 2026First reported | |
M10 · Dangerous capabilities and risk determinations 16 values
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentClaude Opus 4.7 System Card · 16 Apr 2026 | ||
| AI R&D kernel taskBest speedup | Claude Opus 4.7 System Card16 Apr 2026 | |
| AI R&D LLM training taskAverage speedup | Claude Opus 4.7 System Card16 Apr 2026 | |
| AI R&D novel compiler taskPass rate | Claude Opus 4.7 System Card16 Apr 2026 | |
| AI R&D quadruped RL taskHighest score | Claude Opus 4.7 System Card16 Apr 2026 | |
| Autonomy threat model 2 (automated R&D)RSP determination | Claude Opus 4.7 System Card16 Apr 2026 | |
| CB-1 threat modelRSP determination | Claude Opus 4.7 System Card16 Apr 2026 | |
| CB-2 threat modelRSP determination | Claude Opus 4.7 System Card16 Apr 2026 | |
| CybenchSuccess rate, pass@1Condition: 35 challenges | Claude Opus 4.7 System Card16 Apr 2026 | |
| CyberGymTargeted vuln reproduction rate | Claude Opus 4.7 System Card16 Apr 2026 | |
| DNA synthesis screening evasionPathogens with viable evasionCondition: of 10 pathogens | Claude Opus 4.7 System Card16 Apr 2026 | |
| Long-form virology task 1Score | Claude Opus 4.7 System Card16 Apr 2026 | |
| Long-form virology task 2Score | Claude Opus 4.7 System Card16 Apr 2026 | |
| Multimodal virology (VCT)Score | Claude Opus 4.7 System Card16 Apr 2026 | |
| From other documents | ||
| Anthropic ECI (AECI)Capability indexClaude Opus 4.8 System Card28 May 2026First reported | Claude Opus 4.8 System Card28 May 2026First reported | |
| CyberGymTargeted vuln reproduction rateClaude Opus 4.8 System Card28 May 2026First reported | Claude Opus 4.8 System Card28 May 2026First reported | |
| AI R&D LLM training taskAverage speedupCondition: easy variantClaude Opus 5 System Card24 Jul 2026First reported | Claude Opus 5 System Card24 Jul 2026First reported | |
M12 · Chain-of-thought monitorability 1 value
| Evaluation and metric | Value | Document |
|---|---|---|
| From its own documentClaude Opus 4.7 System Card · 16 Apr 2026 | ||
| Accidental CoT supervision in RLShare of training episodes affected | Claude Opus 4.7 System Card16 Apr 2026 | |
Risk determinations
The developer's formal decisions about Claude Opus 4.7 under its framework, as printed. Levels from different frameworks do not map onto one another.
| Domain | Level as printed | Framework and document |
|---|---|---|
| Bio/chem | applicable code, not the printed wording Responsible Scaling Policy Claude Opus 4.7 System Card · 16 Apr 2026 | Responsible Scaling Policy Claude Opus 4.7 System Card · 16 Apr 2026 |
| Bio/chem | not applicable Responsible Scaling Policy Claude Opus 4.7 System Card · 16 Apr 2026 | Responsible Scaling Policy Claude Opus 4.7 System Card · 16 Apr 2026 |
| AI R&D / autonomy | not applicable Responsible Scaling Policy Claude Opus 4.7 System Card · 16 Apr 2026 | Responsible Scaling Policy Claude Opus 4.7 System Card · 16 Apr 2026 |
“Code, not the printed wording”: the version 0 file stored a code for this level rather than the words the document printed. The wording will be read again from the source.