feat: add direct xAI Grok support (stack 4/4) - #3068
Draft
antonvishal wants to merge 45 commits into
Draft
antonvishal wants to merge 45 commits into
antonvishal wants to merge 45 commits into
Conversation
# Conflicts: # packages/sdk-go/internal/extensionassets/stagehand-extension.zip
🦋 Changeset detectedLatest commit: 1c1740a The changes in this PR will be included in the next version bump. This PR includes changesets to release 25 packages
Not sure what this means? Click here to learn what changesets are. Click here if you're a maintainer who wants to add another changeset to this PR |
Contributor
|
This PR is from an external contributor and must be approved by a stagehand team member with write access before CI can run. |
github-actions
Bot
added
external-contributor
|
Review the following changes in direct dependencies. Learn more about Socket for GitHub.
|
# Conflicts: # packages/sdk-go/internal/extensionassets/stagehand-extension.zip # pnpm-lock.yaml
…2-pr-2921 # Conflicts: # packages/sdk-go/internal/extensionassets/stagehand-extension.zip
This was referenced Oct 3, 2026
antonvishal
force-pushed
the
model-allowlist
branch
from
October 3, 2026 17:44
270a0a5 to
f54173e
Compare
antonvishal
changed the title
Use current OpenAI, Anthropic, and Google catalog IDs in SDK examples, generated schema samples, eval defaults, and harness pricing. GPT-6 models default to hybrid agent mode. CUA defaults stay on the existing computer-use IDs.
Point v4 examples and the models page at current OpenAI, Anthropic, and Google IDs, and rank recommendations from Browserbase Benchmark v2.
# Conflicts: # packages/docs/v4/configuration/models.mdx
Point xAI docs, tests, and eval pricing at the current Grok 4.7 API ID.
Keep oxfmt happy after the current provider ID strings.
Leave eval harness IDs unchanged. Use gpt-6-luna for former mini defaults, keep Sol for larger OpenAI fixtures.
Keep Sol for default examples, Astra and Fable as the largest models, and leave eval harness IDs alone.
Keep the eval pricing table unchanged while docs and unit tests use grok-4.7.
Update only Sol and Luna IDs in the recommendations table.
Use opus-5-5 for accuracy, gpt-6-sol for speed, and keep Sol/Luna for production and cost.
# Conflicts: # pnpm-lock.yaml # pnpm-workspace.yaml
…ion-protocol-sdks Co-authored-by: Cursor <cursoragent@cursor.com> # Conflicts: # pnpm-lock.yaml # pnpm-workspace.yaml
…er-docs Co-authored-by: Cursor <cursoragent@cursor.com> # Conflicts: # pnpm-lock.yaml # pnpm-workspace.yaml
…-models-extension-protocol-sdks
…sdks' into vishal/model-provider-docs # Conflicts: # pnpm-lock.yaml
Drop unused Groq/Cerebras catalog pins that leaked in from an independent main merge.
# Conflicts: # packages/sdk-go/internal/extensionassets/stagehand-extension.zip
# Conflicts: # packages/sdk-go/internal/extensionassets/stagehand-extension.zip
# Conflicts: # packages/sdk-go/internal/extensionassets/stagehand-extension.zip
…r3098-conflict-fix
…odel-allowlist-conflict-fix
This branch has not been deployed
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Why
Stack 4/4. Add direct xAI Grok inference after provider-only validation and its documentation land.
Merge order: #2921, #2941, #3098, #3068. Fork PRs target upstream
main; merge in this order. Each branch contains its predecessor. Depends on #3098. This is the final layer of the original draft.What
Add
xai/model names and the AI SDK xAI Responses provider for direct and agent inference. Require a provider API key because Model Gateway does not support xAI. Reject unsupported stop sequences. Docs and unit tests usexai/grok-4.7. Eval pricing and eval harness IDs are unchanged.Focused diff: antonvishal/stagehand@vishal/model-provider-docs...model-allowlist
How
53 xAI-related extension tests, 228 Python SDK tests, and Go SDK tests passed.
Reviewer: verify xAI provider construction, tool and JSON-schema request serialization, unchanged upstream errors, and keyless/stop-sequence rejection. Tests intercept requests; they do not run paid live xAI inference.
ModelProvidergainsxai, so exhaustive consumers may need changes.