Improvements to base openai compatible #9462
Conversation
dosubot
bot
added
size:L
Re-reviewed commit b84d7d1 (merge from main). The new changes look good - the tool call fallback mechanism is correctly implemented across OpenAI, OpenRouter, and Roo providers. However, the previously flagged issue remains unaddressed.
Mention @roomote in a comment to request specific changes to this pull request or fix all unresolved issues. |
![roomote[bot]](https://avatars.githubusercontent.com/in/1546624?s=60&v=4){kind=small}
| cacheWriteTokens: cacheWriteTokens || undefined, | ||
| cacheReadTokens: cacheReadTokens || undefined, | ||
| totalCost, | ||
| } |
There was a problem hiding this comment.
The new processUsageMetrics method in the base class now includes cost calculation via calculateApiCostOpenAI. However, FeatherlessHandler has a custom createMessage implementation for DeepSeek-R1 models that yields usage without cost calculation (lines 79-85). This creates an inconsistency: non-R1 models (which use super.createMessage()) will get cost calculation, but R1 models won't. Consider updating the R1 path to also use processUsageMetrics or call calculateApiCostOpenAI directly to maintain consistent cost tracking across all Featherless models.
Fix it with Roo Code or mention @roomote and request a fix.
hannesrudolph
added
the
Issue/PR - Triage
|
@roomote please resolve the conflict with origin/main |
Resolved merge conflicts with origin/main. All tests passed. |
Ideally we can share more between providers - I think this works for Groq and Baseten at least.
Important
Enhance base OpenAI-compatible provider for better compatibility and flexibility across different providers, with updates to test assertions and usage metrics handling.
toMatchObjectfor partial matching inbase-openai-compatible-provider.spec.ts,featherless.spec.ts,fireworks.spec.ts,groq.spec.ts,io-intelligence.spec.ts,sambanova.spec.ts, andzai.spec.ts.createMessageandcompletePromptmethods inbase-openai-compatible-provider.tsto handle reasoning content and usage metrics more flexibly.processUsageMetricsmethod inbase-openai-compatible-provider.tsto calculate and yield usage metrics.createStreamandcreateMessagemethods to include reasoning parameters if supported by the model.groq.tsandzai.tsby leveraging the base provider's functionality.cacheWriteTokensandcacheReadTokensare undefined when zero ingroq.spec.tsandzai.spec.ts.This description was created by
for 3b468dd. You can customize this summary. It will automatically update as commits are pushed.