Sign in
← AI PulseAug 21, 2026

Deep · news · Multi-source brief

Vercel AI SDK Updates Include DeepSeek V4 Flash Vision Exp and Gateway Webhook Support

Vercel has released updates to its AI SDK, introducing support for DeepSeek V4 Flash Vision Exp image input and Files API, alongside webhook functionality for video models in the gateway.

By Illumora Editorial

Source · Aug 21, 2026, 3:45 PM · On Illumora · Aug 21, 2026, 3:58 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/svelte@5.0.74 · vercel/ai
Save

Vercel's AI SDK has received updates across several packages, including @ai-sdk/svelte@5.0.74, @ai-sdk/workflow@2.0.4, @ai-sdk/tui@1.0.75, and @ai-sdk/vue@4.0.74. These releases, managed by the Changesets release GitHub action, incorporate new features for multimodal capabilities and gateway interactions.

Key Points

  • The @ai-sdk/deepseek@3.0.30 package now supports DeepSeek V4 Flash Vision Exp for image input.
  • Files API support has been added for DeepSeek V4 Flash Vision Exp within @ai-sdk/deepseek@3.0.30.
  • The @ai-sdk/gateway@4.0.60 package implements a handleWebhookOption for video models.
  • This webhook implementation allows generateVideo({ webhook }) to register a factory URL as the gateway's callbackUrl.
  • The gateway will now await delivery via webhook instead of reverting to polling for video generation.
  • The core ai package has been updated to version 7.0.74 across these releases.

Context

According to the Vercel AI SDK Changelog, these updates introduce specific functionalities for developers working with AI models. The addition of DeepSeek V4 Flash Vision Exp image input and Files API support expands the multimodal capabilities available through the SDK. Furthermore, the handleWebhookOption in the gateway for video models streamlines the process of receiving video generation results, moving from a polling mechanism to a more efficient webhook-based delivery system.

Why It Matters

These updates offer builders enhanced options for integrating advanced multimodal models and improve the efficiency of asynchronous operations. The shift to webhook-based delivery for video models can reduce latency and resource consumption associated with continuous polling, while DeepSeek V4 Flash Vision Exp support broadens the range of vision-enabled applications developers can create.

What To Do

  • Review the @ai-sdk/deepseek package documentation for details on implementing DeepSeek V4 Flash Vision Exp image input and Files API.
  • Examine the @ai-sdk/gateway documentation to understand the new handleWebhookOption for video models and its callbackUrl registration.
  • Consider updating to ai@7.0.74 and related packages to leverage these new features.
  • Test the generateVideo({ webhook }) functionality to evaluate its impact on delivery workflows.