llm-anthropic 0.25.1 Adds Claude Opus 4.8 and Fast Mode

The 0.25.1 release of the llm‑anthropic Python library introduces the Claude Opus 4.8 model, available under the identifier claude‑opus‑4.8. It also adds a new –o fast 1 command‑line option that enables fast mode for accounts with the feature enabled. The default max_tokens for each model now defaults to the model’s maximum output instead of the previous hard‑coded 8,192.

These changes improve developer experience by providing a higher‑performance model and simplifying token management. The release notes recommend upgrading to 0.25.1 to take advantage of the new model and fast mode, and to adjust code that relied on the old max_tokens default.