tllama

The Ollama API, reimplemented in Rust, backed by T3 Chat's models. Wine for Ollama clients.

status: Experiment
role: Builder
when: 2025
kind: AI tooling

tllama

AI tooling2025

ansh.bhalala.org/projects/tllama

the TSPL2 that prints this label

SIZE 101.6 mm, 50.8 mm
GAP 3 mm, 0 mm
DIRECTION 1
CLS
TEXT 32,28,"4",0,1,1,"tllama"
TEXT 32,92,"2",0,1,1,"AI tooling"
TEXT 32,124,"2",0,1,1,"2025"
BARCODE 32,196,"128",120,0,0,1,2,"https://ansh.bhalala.org/projects/tllama"
PRINT 1,1

tllama replicates the Ollama API in Rust and answers it with T3 Chat's models. Any tool that already speaks Ollama works unchanged. I think of it as Wine for Ollama clients.

How it works

An Axum service implements /api/generate, /api/chat, and /api/tags, including streaming, and maps each request onto the upstream model. It runs in Docker.

The interesting part is the adapter: making one provider look exactly like another, down to the streaming format, so client code never has to know.

Built with

Rust
Axum
Docker