Skip to content

Conversation

@shuv1337
Copy link
Contributor

@shuv1337 shuv1337 commented Jan 4, 2026

Added --variant CLI option to opencode run command to specify provider-specific model reasoning effort levels (e.g., high, max, minimal). The implementation passes variant through both command and prompt execution paths, storing it in user messages where it's later merged into model options during LLM streaming. Includes comprehensive unit tests covering both execution paths and backwards compatibility.

closes: #6706

shuv1337 and others added 4 commits January 3, 2026 20:29
Add --variant flag to specify model variant (provider-specific reasoning
effort like high, max, minimal). The variant is passed to both
sdk.session.command() and sdk.session.prompt() calls.
Test that variant is correctly passed to sdk.session.prompt() and
sdk.session.command() when using the --variant flag.
@rekram1-node rekram1-node changed the title Feat/cli run variant option feat: add --variant flag to run command Jan 4, 2026
@rekram1-node rekram1-node merged commit 72cef0d into anomalyco:dev Jan 4, 2026
2 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

[FEATURE]: cannot use GPT-5.2 high in CLI mode (aka opencode run)

2 participants