Vercel Developers
@vercel_dev
Grok Build 0.1 is available on Vercel AI Gateway. xAI's beta coding model in Grok Build CLI. ๐๐๐๐๐: '๐ก๐๐/๐๐๐๐-๐๐๐๐๐-๐ถ.๐ท' https://t.co/sqUO2Bsvbp
8likes
View on XVercel Brings Grok Build 0.1 to AI Gateway for Custom Agentic Workflows
Vercelยท Updated
Vercel added xAI's Grok Build 0.1 beta coding model to its AI Gateway, allowing developers to call the specialized reasoning model through the AI SDK. This integration enables teams to build custom autonomous agents using the same weights that power xAI's terminal-based engineering tools.
Vercel, a frontend cloud platform and creator of the AI SDK, integrated Grok Build 0.1 into its AI Gateway. This beta coding model from xAI is specifically trained for agentic coding (AI that autonomously navigates and edits codebases). It is a reasoning-only model with a non-configurable reasoning budget.
- Model identifier
- xai/grok-build-0.1
- Specialization
- Agentic coding
- Reasoning mode
- Always-on (non-configurable)
- Availability
- Vercel AI Gateway
- Pricing
- No markup (Bring Your Own Key)
This integration follows the recent launch of Grok Build, xAI's terminal-based agentic CLI. By bringing the underlying model to the AI Gateway, Vercel allows developers to move beyond xAI's pre-built tools. It reinforces a trend where agentic workloads drive production traffic, requiring infrastructure that handles high-volume autonomous reasoning.
You can now access the model using the xai/grok-build-0.1 identifier within the AI SDK. The AI Gateway provides unified usage tracking, cost monitoring, and automatic provider retries to ensure uptime for long-running agent sessions. It is available now with no price markup, provided you use your own xAI API key.
Still wondering? A few quick answers below.
Grok Build 0.1 is a beta coding model from xAI specifically trained for agentic coding tasks. Unlike general-purpose models, it is designed to power autonomous engineering workflows where an AI agent navigates codebases and performs multi-step edits. It operates as a reasoning model and does not offer a non-reasoning or fast mode.
You can access the model through the Vercel AI Gateway by using the model identifier xai/grok-build-0.1 within the Vercel AI SDK. This allows you to call the model directly in your application code for streaming text or generating responses while leveraging Vercel's infrastructure for usage tracking, observability, and automatic retries.
The AI Gateway provides a unified API that simplifies calling models from different providers. It offers built-in custom reporting, cost tracking, and performance optimizations like intelligent provider routing and failover. These features help ensure higher uptime and better reliability for production applications compared to calling a single provider's API directly.
The model is currently in early access as a beta release. Developers can use it on the Vercel AI Gateway with no additional price markup from Vercel. However, it requires a Bring Your Own Key configuration, meaning you must have your own xAI account and API key to authenticate the requests.
No, the reasoning effort for this specific model is not configurable. While some other reasoning models allow developers to choose between different levels of internal deliberation to balance speed and accuracy, Grok Build 0.1 always operates with its default reasoning process enabled and does not have a separate non-reasoning mode.
Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards โ
