Vercel AI Gateway adds MiniMax H3 video model

// 1h agoPRODUCT UPDATE

Vercel AI Gateway adds MiniMax H3 video model

Vercel announced the addition of MiniMax H3 to the Vercel AI Gateway, enabling developers to generate AI videos directly within their applications. By using the `generateVideo` function in Vercel's AI SDK with the `minimax/minimax-h3` model identifier, developers can programmatically synthesize video content using standard text prompts.

// ANALYSIS

Bringing MiniMax H3 to Vercel AI Gateway lowers the barrier for developers building multimodal applications by offering video generation through a single, managed SDK interface.

* Simplifies developer access to video synthesis models alongside text and image generation.

* Expands Vercel's multimodal AI ecosystem with alternative model providers.

* Eliminates complex API integrations by standardizing access via Vercel's unified AI Gateway.

// TAGS

vercelai-gatewayminimaxminimax-h3video-genai-sdk

DISCOVERED

1h ago

2026-07-31

PUBLISHED

2h ago

2026-07-31

RELEVANCE

7/ 10

AUTHOR

vercel_dev

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

UPDATE1h ago

Open Science v0.8.1 enhances long-running research workflows

AIPOCH has released Open Science v0.8.1, an update dedicated to enhancing long-running research workflows, workspace management, and artifact reliability. As AI-driven research projects increase in scope and complexity, this release provides researchers with improved visibility into context usage and file management to ensure consistent, reproducible results.

MODEL1h ago

Hailuo AI H3 lands on OpenRouter

OpenRouter has integrated MiniMax's open-weights Hailuo AI H3 video model, supporting multimodal inputs to generate native audio-visual outputs. Accessible via OpenRouter's Video API, H3 features first and last frame conditioning, up to 12 reference files, and instruction-led video editing.

OPEN SOURCE1h ago

Quill ships crash-resilient local meeting recording

Quill is an open-source macOS app designed for private, crash-resilient local meeting recording and transcription. Using Core Audio Format (CAF) containers and a filesystem job queue, it preserves raw audio during crashes and automatically recovers interrupted transcription tasks upon restart.