← Back to Vault

Vision Integration via MCP

Tom Spencer · Category: frameworks_and_exercises

Use an MCP server to intercept image uploads from GPT OSS models and route them to Qwen-2.5-VL for flexible image description and OCR within a unified pipeline.