Sign in
← AI PulseAug 3, 2026

Wire · news · Multi-source brief

Vercel AI SDK Adds Asynchronous Video Generation APIs

The Vercel AI SDK has introduced asynchronous APIs for its experimental video model interface, allowing for polling and webhook-based orchestration of video generation.

By Illumora Editorial

Source · Aug 3, 2026, 11:29 PM · On Illumora · Aug 3, 2026, 11:32 PM

Media from the primary source — shown here so you can stay on Illumora.

Synthesized from multiple allowlisted primaries on the same event. Lanes →

Brief drafted by Illumora’s editorial model from the linked primary source. Ops desk reviews flagged pieces. How we write →

Read the source →Vercel AI SDK Changelog — Release @ai-sdk/xai@4.0.26 · vercel/ai
Save

Vercel AI SDK, in its @ai-sdk/xai@4.0.26 release, has added asynchronous APIs for the VideoModelV4 interface. This update, released on August 3, 2024, introduces a start/status flow for video generation, enabling models to implement doStart, doStatus, and handleWebhookOption.

These new capabilities allow the experimental_generateVideo function to accept poll and webhook options, facilitating completion orchestration through either polling or webhooks. Polling configurations can also utilize a custom delay implementation for compatibility with durable workflows.

Key Points

  • The @ai-sdk/xai@4.0.26 release introduces asynchronous APIs for video generation.
  • The VideoModelV4 interface now supports doStart, doStatus, and handleWebhookOption.
  • experimental_generateVideo accepts poll and webhook options for completion orchestration.
  • Polling configurations can include a custom delay implementation for durable workflow compatibility.
  • This update was released on August 3, 2024.
  • The @ai-sdk/xai@3.0.114 release, also on August 3, 2024, addressed an issue where video generation would hang during status polling.

Context

According to the Vercel AI SDK Changelog, the asynchronous APIs enhance the experimental video model interface. A previous release, @ai-sdk/xai@3.0.114, fixed a bug where video generation would hang while polling status, indicating ongoing development and refinement of video generation capabilities within the SDK.

Why It Matters

This update provides developers with more flexible and robust methods for integrating video generation into applications, particularly for long-running processes that benefit from asynchronous handling and status monitoring.

What To Do

  • Review the @ai-sdk/xai@4.0.26 release notes for detailed implementation specifics.
  • Explore the doStart, doStatus, and handleWebhookOption methods for custom video model implementations.
  • Test the poll and webhook options with experimental_generateVideo for asynchronous workflows.
  • Note the custom delay implementation for polling, especially for durable workflow scenarios.