fix(openrouter): correct reserved tokens for google/gemini-2.5-pro-preview #3896
Conversation
…eview to 66k and add test coverage
![ellipsis-dev[bot]](https://avatars.githubusercontent.com/in/64358?s=60&v=4){kind=small}
| @@ -106,7 +106,11 @@ export async function getOpenRouterModels(options?: ApiHandlerOptions): Promise< | |||
| id, | |||
| model, | |||
| modality: architecture?.modality, | |||
| maxTokens: id.startsWith("anthropic/") ? top_provider?.max_completion_tokens : 0, | |||
| maxTokens: id.startsWith("anthropic/") | |||
There was a problem hiding this comment.
The nested ternary for maxTokens reduces clarity. Consider extracting the condition (and using a named constant for 66000) to improve maintainability.
| maxTokens: id.startsWith("anthropic/") ? top_provider?.max_completion_tokens : 0, | ||
| maxTokens: id.startsWith("anthropic/") | ||
| ? top_provider?.max_completion_tokens | ||
| : id === "google/gemini-2.5-pro-preview" |
There was a problem hiding this comment.
Does openrouter just not include this number? Is that a bug?
There was a problem hiding this comment.
after doing some debugging ive hit a wall, the model file we fetch from openrouter does have it as 66k but without this i see 209k and hardcoding is showing 66k correctly. am still trying to figure it out why its happening with gemini or its an issue for all models in openrouter
|
@mrubens its an issue with flash thinking also , i will test some more models tomorrow to see if its just gemini issue or entire openrouter model list |
Possible fix for #3817
Important
Sets
maxTokensto 66000 forgoogle/gemini-2.5-pro-previewingetOpenRouterModelsand adds a test for it.maxTokensto 66000 forgoogle/gemini-2.5-pro-previewingetOpenRouterModelsinopenrouter.ts.openrouter.test.tsto verifymaxTokensis set to 66000 forgoogle/gemini-2.5-pro-preview.This description was created by
for cbf0fe3. You can customize this summary. It will automatically update as commits are pushed.