Brief
Claude Code v2.1.267 Adds Effort Cap and Fresh System Prompt Option
The release introduces a maxEffortLevel setting that caps effort on every provider, and a flag to render the system prompt fresh on each request.
Claude Code v2.1.267 adds a maxEffortLevel setting, which can be set at the top level or per model under modelSettings. It caps the effort level on every provider, including Bedrock, Vertex and Foundry, while users can still pick a lower level. The release also adds --system-prompt-snapshot off to render the system prompt fresh on every request instead of reusing the conversation's recorded prompt, intended for iterating on prompt text.
Our reading
Our reading is that this gives teams a way to enforce a ceiling on compute effort across providers without removing user choice to go lower.
What to do or watch
If you administer Claude Code across multiple providers, decide whether maxEffortLevel belongs at the top level or per model under modelSettings, keeping in mind it sets a ceiling on Bedrock, Vertex and Foundry while users can still choose lower. Watch what --system-prompt-snapshot off changes in practice when rendering the system prompt fresh on every request, and note the source text does not say what that costs or how it interacts with prompt reuse.
Source details and supporting facts
Each line is stated by the page named above it.
Stated by github.com
- Added maxEffortLevel setting (top-level or per model under modelSettings): caps the effort level on every provider, including Bedrock, Vertex and Foundry; users can still pick a lower level
- Added --system-prompt-snapshot off to render the system prompt fresh on every request instead of reusing the conversation's recorded prompt (for iterating on prompt text)
Sources
- Anthropic Claude Code release notes (GitHub)Text stored 15 September 2026
How this story was checked. Written from the 1 page listed above, stored 15 September 2026; claims checked against that stored text on 15 September 2026.
What that means
- 2 of 2 reported statements were confirmed against the page that carries them; the rest were removed rather than published.
- Figures in the text were required to appear in the stored source text: yes. Identifiers: yes.
- The check reads stored text only: no claim rests on a fresh look that did not happen.
- Where the reporting was silent, the text says so instead of filling the gap.