Seedance 2.5 hits Vercel AI Gateway

// 46d agoPRODUCT UPDATE

Seedance 2.5 hits Vercel AI Gateway

Vercel has announced the integration of Bytedance's Seedance 2.5 model into its AI Gateway ecosystem. Developers can now programmatically generate videos using this model through a simple generateVideo function call, passing prompts like "dreamcore san francisco".

// ANALYSIS

This integration significantly broadens the multimodal capabilities accessible directly through Vercel's standard tooling. It seamlessly incorporates advanced AI video generation into existing Vercel AI SDK workflows and reduces the friction of setting up separate video generation APIs by leveraging the unified AI Gateway. This underscores a growing industry trend of unified platforms offering diverse generative models within a single developer experience.

// TAGS

aivideo-genvercelai-gatewayseedance-2.5bytedance

DISCOVERED

46d ago

2026-08-09

PUBLISHED

46d ago

2026-08-09

RELEVANCE

6/ 10

AUTHOR

vercel_dev

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

RESEARCH38m ago

Pokee AI whitepaper maps enterprise agent security

Pokee AI collaborated with researchers from UIUC and the University of Chicago to publish a 36-page whitepaper defining a threat landscape and hardened architecture for enterprise AI agents. The paper maps critical vulnerabilities across agent trajectories—including indirect prompt injection and tool composition risks—while introducing benchmark evaluations on Pokee-Isaac 28B that balance attack defense against benign task completion.

UPDATE44m ago

OpenMuse transitions built-in agent to TanStack AI

OpenMuse, an open-source personal AI agent developed in the CopilotKit ecosystem, has transitioned its built-in agent engine to TanStack AI. The integration leverages TanStack AI's framework-agnostic TypeScript toolkit to provide standardized building blocks for durable agent loops, provider adapters, and tool invocation.

OPEN SOURCE56m ago

OpenClaw prunes 400,000 lines of redundant tests

OpenClaw deleted approximately 400,000 lines of redundant code from its test suite without materially impacting test coverage, addressing a growing issue where AI coding models generate low-value tests for minor modifications. To eliminate the bloat, the team open-sourced a dedicated test-audit skill that establishes strict authoring gates and uses mathematically bounded prompts to prune implementation-coupled tests.