nixamp 0.24 keeps every transcript once, under the media it came from, detects the language it is hearing, translates captions with open models on its own CPU, and serves its tools as an MCP at nixamp.com/mcp.
Read the original (By Anthony Ettinger, via Chovy's Blog): https://dev.profullstack.com/~anthony/blog/148-post.html
Hire us: https://profullstack.com/plans
