Start in Console

Turn any repo into a podcast.

Paste a public GitHub repo. Qwen 3.8 Max or Flash writes a two-host episode from the code, two GMI MaaS voices read it, and you get one WAV, with tokens, latency, speed and cost for every call on the right.

or drop a repo folder or .zip here · ·

Paste a key, then a repo URL or a folder.

Read repoREADME + key files

Write scriptQwen3.8-Max

Render voicestwo MaaS voices

Mixstereo WAV

Script model TTS model Host A voice Host B voice Length GitHub token

Episode

Script lands here. Lines turn solid as their audio renders.

15TTS models on GMI MaaS

680+voices, two per episode

262Ktoken context on Qwen3.8-Max

Every model, one API.
GMI Inference Engine.

Script and voices both run through Model as a Service: one key, OpenAI-compatible chat, and a single request queue for audio, image and video models.

Explore Model as a Service Start in Console

Build this on GMI Cloud

One endpoint. 200+ models. Pay per token, scale to zero.

Turn any repo into a podcast.

Episode

Every model, one API.GMI Inference Engine.

Build this on GMI Cloud

Every model, one API.
GMI Inference Engine.