Overview
VLM Run Gateway MCP connects Claude to VLM Run's open-weight visual intelligence platform: VLMs, ViTs, and OCR models in one unified catalog. Read and interpret PDFs, scans, receipts, forms, images, video, and audio, with no infrastructure to manage. Key capabilities - Documents & OCR: PDFs/scans → markdown or structured JSON, with layout, tables, and text-level bounding boxes - Multimodal understanding: captions, visual Q&A, video summaries, and speech transcription - Physical AI: detect and count objects, segment and track regions, estimate body pose and more API Docs: https://docs.vlm.run
