VLM Run Launches Unified API Gateway for 21 Visual Models

spillai · x · 2026-09-01

VLM Run has launched Gateway, a unified OpenAI-compatible API to run various open-weight vision models. It currently supports 21 models (including Qwen, Kimi, Muse) covering OCR, document parsing, video understanding, and image detection. Users can switch models simply by changing a parameter without managing separate endpoints.

Related event: VLM Run Launches Gateway for Unified Access to 21 Vision Models(2 posts)→

Original post →

More from Multimodal

Multimodal channel →