Add xhigh output effort / verbosity for Opus 4.7#2526
Conversation
| low: { reasoning: { enabled: true, effort: 'low' }, verbosity: 'low' }, | ||
| medium: { reasoning: { enabled: true, effort: 'medium' }, verbosity: 'medium' }, | ||
| high: { reasoning: { enabled: true, effort: 'high' }, verbosity: 'high' }, | ||
| xhigh: { reasoning: { enabled: true, effort: 'xhigh' }, verbosity: 'xhigh' }, |
There was a problem hiding this comment.
verbosity is the thing that matters, see: https://openrouter.ai/docs/guides/evaluate-and-optimize/model-migrations/claude-4-7
There was a problem hiding this comment.
this file is not important, it is only used for grok-code-fast-1
| effort: verbosity, | ||
| disableParallelToolUse: request.parallel_tool_calls === false || isLegacyExtension, | ||
| } satisfies AnthropicProviderOptions, | ||
| openai: { |
There was a problem hiding this comment.
WARNING: Removing textVerbosity drops verbosity handling for Martian requests
grokCodeFastOptimizedRequest() routes Martian chat-completions traffic through this helper, and request.verbosity is no longer propagated anywhere in commonParams. After this change those requests still accept a verbosity field, but the downstream Responses call ignores it and falls back to the provider default, which looks like an unrelated regression.
Code Review SummaryStatus: 1 Issue Found | Recommendation: Address before merge Overview
Fix these issues in Kilo Cloud Issue Details (click to expand)WARNING
Other Observations (not in diff)None. Files Reviewed (5 files)
Reviewed by gpt-5.4-20260305 · 849,722 tokens |
https://openrouter.ai/docs/guides/evaluate-and-optimize/model-migrations/claude-4-7