Sign in
← AI PulseAug 6, 2026

Wire · news · Multi-source brief

Vercel AI SDK Adds Fish Audio Provider, Enhances ToolLoopAgent and Instruction Handling

The Vercel AI SDK has introduced a new Fish Audio provider with speech and transcription models, alongside updates to ToolLoopAgent callbacks and instruction middleware.

By Illumora Editorial

Source · Aug 6, 2026, 12:58 AM · On Illumora · Aug 6, 2026, 1:08 AM

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/tui@1.0.55 · vercel/ai
Save

The Vercel AI SDK has released updates, including the addition of a Fish Audio provider. This new provider supports both speech and transcription models, as detailed in the changelog for @ai-sdk/fish-audio@3.0.0.

Key Points

  • The Fish Audio provider has been added to the Vercel AI SDK, supporting speech and transcription models.
  • ToolLoopAgent prepareCall callbacks can now read and override the top-level reasoning option.
  • A defaultInstructionsMiddleware has been introduced for applying default language model instructions while preserving call-level overrides.
  • Preceding assistant messages are now preserved when regenerating a response.
  • The SDK rejects spliced Anthropic generations but allows duplicate message start events for the active message.

Context

According to the Vercel AI SDK changelog, these updates are part of a series of package releases, including @ai-sdk/tui@1.0.55, @ai-sdk/vercel@3.0.24, @ai-sdk/valibot@3.0.22, and @ai-sdk/voyage@2.0.22. The changes reflect ongoing development in the ai@7.0.54 package and its dependencies, such as @ai-sdk/amazon-bedrock@5.0.45 and @ai-sdk/anthropic@4.0.32.

Why It Matters

The addition of the Fish Audio provider expands the SDK's multimodal capabilities, offering developers new options for integrating speech and transcription functionalities. Enhancements to ToolLoopAgent and instruction handling provide greater control and flexibility for managing language model interactions and agent behavior.

What To Do

  • Review the updated documentation for the Fish Audio provider to understand its speech and transcription model capabilities.
  • Examine the ToolLoopAgent prepareCall callbacks for opportunities to customize reasoning options.
  • Investigate the defaultInstructionsMiddleware to implement consistent language model instructions with override capabilities.
  • Note the behavior regarding Anthropic generations, specifically the rejection of spliced generations and allowance of duplicate message start events.