On 24 July 2026, Anthropic announced Claude Opus 5. Anthropic targeted near-Fable capability at half its price while improving verification, computer use and professional artifact work.
This release brief was checked against first-party material on 10 August 2026. The date above is the public announcement date, not the date a repository was created or a third-party provider added the model. Where access or weights arrived later, that distinction is recorded below.
Release record
| Field | Verified detail |
|---|---|
| Announcement | 24 July 2026 |
| Availability or weight release | 24 July 2026 on all Claude platforms and the Anthropic API |
| Release type | proprietary frontier general-purpose model |
| Access | Claude products and API at $5 input and $25 output per million tokens |
| Architecture | an undisclosed Opus model with effort controls, automatic fallback and optional fast mode |
| Maximum stated context | the Opus 5 limit documented by Anthropic |
What changed
Opus 5 is designed as a practical default rather than the most restricted Anthropic model. It improves across science and coding while retaining Opus 4.8 pricing, and Anthropic says its safeguard classifiers intervene about 85% less often than Fable 5’s. The launch also adds automatic model fallbacks, which means an evaluation must log whether Opus 5 or a fallback produced the final result.
The practical comparison is therefore not simply whether Claude Opus 5 has the largest headline score. Teams need to ask whether its architecture, access terms, latency, tool behaviour and evaluation setup match the workload they actually intend to run. A model can lead one harness while losing on cost, refusal behaviour, multilingual quality or repeatability in another.
Benchmarks worth retaining
| Evaluation | Reported result | How to read it |
|---|---|---|
| Alignment audit | 2.3 overall misaligned behaviour | Lowest among the recent Claude models in Anthropic’s audit |
| Organic-chemistry internal benchmark | +10.2 points versus Opus 4.8 | Scientific-research improvement |
| Protein-variation task | +7.7 points versus Opus 4.8 | Life-science evaluation gain |
These are release-time results, not independently reproduced guarantees. Many launch charts are effort-versus-cost curves, and Anthropic’s customer evidence spans private benchmarks with different tools and task definitions. Scores should remain attached to the disclosed effort setting, agent harness, tool access, timeout, context-management policy and judge model. Moving a number into a procurement sheet without those conditions creates false comparability.
Architecture and access
Claude Opus 5 is described as an undisclosed Opus model with effort controls, automatic fallback and optional fast mode with the Opus 5 limit documented by Anthropic of stated context. Its access position at verification time is Claude products and API at $5 input and $25 output per million tokens. That wording matters: open weights, source-available weights, an API, a product preview and a research demonstration give adopters very different rights and different levels of reproducibility.
Before deployment, record the exact model identifier or checkpoint, inference stack, quantisation, reasoning setting, region, price schedule and supplier terms. If the release uses a custom licence, read the licence itself rather than relying on the word “open” in launch copy. If it is API-only, preserve the dated documentation and change-notice route because the served snapshot can change without a downloadable artefact.
What an evaluation should test next
For Claude Opus 5, a credible internal gate should include:
- a frozen set of representative tasks with pass, fail and abstain criteria;
- a matched baseline using the same tools, timeout, prompt budget and reviewer rubric;
- repeated runs to expose variance rather than reporting a single best attempt;
- latency, token use and total task cost alongside task success;
- adversarial, multilingual and long-context cases relevant to the real deployment; and
- rollback evidence showing the previous model can be restored safely.
The wider model change-control guide explains how to keep model, prompt, tool and corpus changes reconstructable. The AI dependency inventory guide covers the release and supplier records needed after deployment.
AIEngine verdict
Opus 5 is a major everyday frontier release and the clearest upgrade path from Opus 4.8. Migration should still be gated on fallback visibility, output consistency and real task cost.
This is a launch assessment, not a certification. Benchmark leadership is useful evidence of where to test; it is not authorization to place the model in a high-impact workflow without domain evaluation, security review and an accountable owner.
Primary sources
Image provenance
Hero image: Anthropic official release artwork. The locally served WebP is a crop of the first-party release or model-card asset recorded in the repository provenance manifest.
