feat(ollama): one container, several models, each its own priced app - #62

Open

rickstaa wants to merge 10 commits into

mainfrom

rs/ollama-multi-model

Commits on Aug 8, 2026

fix(vllm): pin the model so the app id cannot lie

rickstaa
and
claude
committed
d394b0c
feat(vllm): register single-shot so the call pays as it runs

rickstaa
and
claude
committed
2a36d0b
fix(vllm): forward every verb and give a generation time to finish

rickstaa
and
claude
committed
de30cc9
docs: note the metadata field under Registration

rickstaa
and
claude
committed
dfba82d
feat(streaming-asr): add the WebSocket example

rickstaa
and
claude
committed
b2d461d
refactor(streaming-asr): fix the model instead of exposing it

rickstaa
and
claude
committed
1f4a49b

Commits on Aug 10, 2026

refactor(realtime-transcription): name it for the capability, pin turbo

rickstaa
and
claude
committed
a43f71c
feat(ollama): one container, several models, each its own priced app

rickstaa
and
claude
committed
d6a4157
refactor(ollama): use the model name as the app id, drop metadata

rickstaa
and
claude
committed
15230d9
fix(ollama): drop a vllm leftover and actually pass MODELS through

rickstaa
and
claude
committed
5bd3d85