Anthropic
Claude Opus 5
Made by Anthropic. Released 2026-07-24. Accessibility not recorded. Cheapest standard rate we hold: $25 per million output tokens, no context split.
What it costs
| Tier | Kind | In | Out | Cached in | From |
|---|---|---|---|---|---|
| batch | text | $2.50 | $13 | — | seller |
| standard | text | $5 | $25 | $0.50 | seller |
per million tokens, US dollars. * marks a rate from a single registry or two that disagree — the best available number rather than a confirmed one. Every other row is confirmed against a seller or a licensed index.
What it scores, by house
| Benchmark | House | Score | Evaluated | Scaffold |
|---|---|---|---|---|
| chess puzzles | Epoch AI | 0.2±0.04 | 2026-08-06 | low |
| chess puzzles | Epoch AI | 0.3±0.05 | 2026-08-06 | — |
| chess puzzles | Epoch AI | 0.4±0.05 | 2026-07-24 | max |
| ebr bench | Epoch AI | 0.5±0.04 | 2026-09-05 | max |
| frontiermath tier 4 v2 | Epoch AI | 0.7±0.07 | 2026-07-24 | max |
| frontiermath tiers 1 3 v2 | Epoch AI | 0.9±0.02 | 2026-07-24 | max |
| gpqa diamond | Epoch AI | 0.9±0.02 | 2026-08-06 | low |
| gpqa diamond | Epoch AI | 0.9±0.02 | 2026-08-06 | — |
| gpqa diamond | Epoch AI | 0.9±0.01 | 2026-07-24 | max |
| mystery game puzzles | Epoch AI | 0.4±0.05 | 2026-08-06 | — |
| mystery game puzzles | Epoch AI | 0.6±0.05 | 2026-07-25 | max |
| otis mock aime 2024 2025near its ceiling | Epoch AI | 1.0±0.02 | 2026-08-06 | — |
| otis mock aime 2024 2025near its ceiling | Epoch AI | 0.9±0.04 | 2026-08-06 | low |
| otis mock aime 2024 2025near its ceiling | Epoch AI | 1.0±0.01 | 2026-07-24 | max |
| simpleqa verified | Epoch AI | 0.6±0.02 | 2026-08-10 | max |
| ale bench | collected by Epoch AI | 2165 | — | high |
| apex agents | collected by Epoch AI | 0.6 | — | max |
| arc aginear its ceiling | collected by Epoch AI | 1.0 | — | high |
| arc aginear its ceiling | collected by Epoch AI | 1.0 | — | max |
| arc agi 2 | collected by Epoch AI | 0.9 | — | max |
| arc agi 2 | collected by Epoch AI | 0.9 | — | high |
| blueprint bench 2 | collected by Epoch AI | 0.3 | — | unknown |
| btf3 | collected by Epoch AI | 0.1 | — | FutureSearch agent |
| critpt | collected by Epoch AI | 0.3 | — | xhigh |
| critpt | collected by Epoch AI | 0.3 | — | medium |
| critpt | collected by Epoch AI | 0.2 | — | low |
| critpt | collected by Epoch AI | 0.3 | — | high |
| critpt | collected by Epoch AI | 0.3 | — | max |
| cursorbench | collected by Epoch AI | 0.7 | — | xhigh |
| cursorbench | collected by Epoch AI | 0.6 | — | medium |
| cursorbench | collected by Epoch AI | 0.7 | — | max |
| cursorbench | collected by Epoch AI | 0.7 | — | high |
| cursorbench | collected by Epoch AI | 0.6 | — | low |
| deepswe | collected by Epoch AI | 0.7 | — | mini-swe-agent |
| dtbenchnear its ceiling | collected by Epoch AI | 1.0 | — | max |
| frontiercode | collected by Epoch AI | 0.5 | — | claude-code |
| gbaeval | collected by Epoch AI | 0.8 | — | unknown |
| gdp pdf | collected by Epoch AI | 0.2 | — | max |
| lmca | collected by Epoch AI | 63.3 | — | max |
| osworld 2 | collected by Epoch AI | 0.2 | — | low |
| osworld 2 | collected by Epoch AI | 0.3 | — | medium |
| osworld 2 | collected by Epoch AI | 0.3 | — | high |
| osworld 2 | collected by Epoch AI | 0.3 | — | xhigh |
| osworld 2 | collected by Epoch AI | 0.3 | — | max |
| posttrainbench | collected by Epoch AI | 0.4 | — | Claude Code |
| proofbench | collected by Epoch AI | 1.0 | — | max |
| scicode | collected by Epoch AI | 0.5 | — | medium |
| scicode | collected by Epoch AI | 0.6 | — | max |
| scicode | collected by Epoch AI | 0.5 | — | low |
| scicode | collected by Epoch AI | 0.5 | — | high |
| scicode | collected by Epoch AI | 0.5 | — | xhigh |
| simplebench | collected by Epoch AI | 0.8 | — | unknown |
| vending bench 2 | collected by Epoch AI | 11182 | — | unknown |
| webdev arena | collected by Epoch AI | 1691 | — | max |
| webdev arena | collected by Epoch AI | 1663 | — | high |
| weirdml | collected by Epoch AI | 0.9 | — | max |
| weirdml | collected by Epoch AI | 0.9 | — | — |
| weirdml | collected by Epoch AI | 0.9 | — | high |
These scores are not comparable between houses and are not added up. A score is dated by when it was evaluated, not when the model was released, and a score without its scaffold is not a measurement — which is why both are printed.
What retires, and when
- deprecation — 2027-07-24 (Active)
Headlines that name it
- Claude Opus 5 deletes production Supabase database after developer grants unrestricted access
- Claude Opus 5 passes 23.9% in new agent construction benchmark
Matched by the model's name appearing in our headline, literally. There is no link from a story to a model in our data, so this is a search result rather than a claim that the story is about it.
Every rate we hold · Why there is no single best model · Everything from Anthropic