@bcherny: Opus 5 is a great model for coding, data analysis, design, biology, knowledge work. More than any of these eval scores,…
Summary
Anthropic's Claude Opus 5 is highlighted as a state-of-the-art model for coding, data analysis, and knowledge work, with unprecedented resistance to prompt injection attacks. The system card reveals that combined defenses reduce prompt injection success rates to near zero.
View Cached Full Text
Cached at: 07/25/26, 06:06 AM
Opus 5 is a great model for coding, data analysis, design, biology, knowledge work.
More than any of these eval scores, what is most exciting to me is something else: Opus 5 is our least prompt injectable model yet. It is a bit buried in the system card, but across PI evals and red teaming, Opus 5 is very hard to prompt inject successfully.
And when layering defenses – strong model alignment, combined with prompt injection probes, combined with Auto Mode in Claude Code – the success rate for prompt injection attacks drops to ~0. This is new and exciting! More about this soon.
https://www-cdn.anthropic.com/c5fbac3f0b1280a933ebd26d3cb8bb9f5bdeaf48/Claude%20Opus%205%20System%20Card.pdf#page=73…
Claude (@claudeai): On several coding and knowledge work evaluations, Opus 5 is the new state-of-the-art:
Similar Articles
Quoting Boris Cherny
Boris Cherny highlights that Opus 5 is the least prompt injectable model yet, based on evaluations and red teaming.
@bcherny: Seeing a number of benchmarks showing Opus is the best model for long-running work. Five tips for running Opus autonomo…
Practical tips for running Anthropic's Claude Opus autonomously for hours or days, such as using auto mode, dynamic workflows, and self-verification; also references the SWE-Marathon benchmark for long-horizon software tasks.
Opus 5's effort dial is not monotonic. Above "high", coding scores go down, and Anthropic's own migration guide says so.
Anthropic's Opus 5 shows non-monotonic performance on coding tasks; the 'high' effort setting outperforms 'max' due to unnecessary refactors. The model also has a 6% higher hallucination rate than Opus 4.8, and safety classifiers may silently fall back to the older model.
@orca_build: Anthropic’s new Opus 4.8 scores 3.6% lower than GPT 5.5 on Terminal-Bench 2.1… …but it’s noticeably better at UI tasks.…
Anthropic's Opus 4.8 scores 3.6% lower than GPT 5.5 on Terminal-Bench 2.1 but excels at UI tasks; Orca's orchestration enables Codex to delegate UI tasks to Claude Code.
@scaling01: Opus 4.8 is the best coding model out there FrontierCode by Cognition is probably the highest quality coding benchmark …
Cognition introduces FrontierCode, a high-quality coding benchmark that goes beyond unit tests to measure code maintainability, regression safety, and quality, with 150 handcrafted tasks by open-source developers.