Skip to content

refactor(inference): select server model by subscription token - #32

Merged
g4vrk merged 1 commit into
g4vrk:masterfrom
jvm-argument:codex/token-model-selection
Sep 26, 2026
Merged

g4vrk merged 1 commit into
g4vrk:masterfrom
jvm-argument:codex/token-model-selection

Conversation

@jvm-argument

Copy link
Copy Markdown
Collaborator

Inference requests no longer send model-family or model-name. Remove both endpoint settings and their paired-value validation from the plugin, and update the English and Russian inference.yml defaults to explain that the server selects the model using the subscription token.

Existing configuration files may retain the obsolete fields; the plugin ignores them. This requires the backend to accept /api/v2/inference/ without model query parameters. The deployed backend now selects and caches the model assignment by subscription token while validating token activity for every request.

Validation: Gradle build passed for common, Paper, and Folia, including Spotless checks.

@g4vrk
g4vrk marked this pull request as ready for review September 26, 2026 22:39
@g4vrk
g4vrk merged commit 96d8d3a into g4vrk:master Sep 26, 2026
1 check 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.

2 participants