Sign in
← AI PulseJul 27, 2026

Wire · news · Multi-source brief

Vercel AI SDK Updates TogetherAI and Mistral Integrations

Vercel has updated its AI SDK, enabling token usage reporting for TogetherAI streaming responses and improving multi-turn conversation reasoning for Mistral models.

By Illumora Editorial

Source · Jul 27, 2026, 9:27 PM · On Illumora · Jul 27, 2026, 9:32 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/togetherai@1.0.50 · vercel/ai
Save

Vercel's AI SDK has received updates for its TogetherAI and Mistral integrations. On July 27, the @ai-sdk/togetherai package was updated to versions 1.0.50 and 2.0.68, enabling streaming responses to report token usage. Concurrently, the @ai-sdk/mistral package was updated to version 3.0.52, addressing an issue to preserve reasoning in multi-turn conversations.

Key Points

  • The @ai-sdk/togetherai package versions 1.0.50 and 2.0.68 now support includeUsage for TogetherAI.
  • This includeUsage feature allows streaming responses from TogetherAI to report token usage.
  • The @ai-sdk/mistral package version 3.0.52 includes a fix for Mistral models.
  • This fix addresses the preservation of reasoning in multi-turn conversations.
  • All updates were released on July 27.

Context

According to Vercel's AI SDK changelog, these updates are patch changes. The TogetherAI update specifically enables a feature for reporting token usage, which is relevant for cost tracking and performance analysis in streaming applications. The Mistral update focuses on improving the consistency and coherence of multi-turn interactions, as stated by Vercel.

Why It Matters

These updates provide developers using the Vercel AI SDK with enhanced functionality and reliability. The ability to track token usage for TogetherAI streaming responses offers greater transparency and control over resource consumption, while the fix for Mistral's multi-turn conversations can lead to more robust and contextually aware applications.

What To Do

  • Developers using @ai-sdk/togetherai should update to version 1.0.50 or 2.0.68 to access token usage reporting.
  • Review the documentation for includeUsage with TogetherAI to understand its implementation.
  • Developers using @ai-sdk/mistral should update to version 3.0.52 to benefit from improved multi-turn conversation reasoning.
  • Test existing applications with the updated packages to ensure compatibility and leverage new features.

Keep Exploring

/atlas/**mistral**-family /atlas/llama-open /techniques/system-user-separation /techniques/multishot