feat: switch xai provider to responses API#18175
feat: switch xai provider to responses API#18175Jaaneek wants to merge 1 commit intoanomalyco:devfrom
Conversation
github-actions
bot
added
needs:compliance
|
Thanks for updating your PR! It now meets our contributing guidelines. 👍 |
Jaaneek
changed the title
| seenToolCalls.add(part.call_id); | ||
| controller.enqueue({ | ||
| type: "tool-input-start", | ||
| \ No newline at end of file |
There was a problem hiding this comment.
ai-sdk/xai responses API only work with the latest versions, those patches align the functionality to the latest version
Ideally we should just bump ai-sdk to v6
There was a problem hiding this comment.
Yeah im planning on doing that in next few days (bumping sdk)
Currently blocked by gitlab provider
There was a problem hiding this comment.
Should we merge this anyway so there is an instant improvement? Im happy to help with the migration if needed
| for (const modelID of Object.keys(provider.models)) { | ||
| if (modelID.includes("multi-agent")) delete provider.models[modelID] | ||
| } |
There was a problem hiding this comment.
not sure if thats the best way to do it but works. We need to hide multi agent until it supports client side tools
There was a problem hiding this comment.
There is an easier way, we just change this to false:
https://github.com/anomalyco/models.dev/blob/123054ae0c1b46cedb536aa0507d94dea48db244/providers/xai/models/grok-4.20-multi-agent-beta-latest.toml#L9
There was a problem hiding this comment.
I updated models.dev, should sync everywhere soon
There was a problem hiding this comment.
thank you, I will remove it from here
I will also open PR to rename the models to drop the "beta" from the name. Unless you can quickly do it.
Jaaneek
force-pushed
the
feat/xai-responses-api
branch
from
615bc30 to
2b8a267
Compare
Jaaneek
force-pushed
the
feat/xai-responses-api
branch
from
2b8a267 to
80d2d37
Compare
Jaaneek
changed the title
2 participants
Issue for this PR
Closes #18179 #3890
Type of change
What does this PR do?
Switches xAI provider from completions API to responses API. The reason for this change is to use 'reasoning.encrypted_content' with reasoning models. This should improve performance in long multi turn conversations.
How did you verify your code works?
I did a lot of manual testing also told agent to test the changes e2e. I tried reading files, writing code, opening images etc..
Screenshots / recordings
Checklist
If you do not follow this template your PR will be automatically rejected.