Ignis
Single-GPU local inference server for Qwen 27B, 125B MoE and image model via OpenAI API.
Not verified yetKnown vulnerabilities
How far pickfit got
pickfit hasn’t verified Ignis yet, so there are no tested install steps.
What follows comes from its README only.
Use it
Before you install: known vulnerabilities
pickfit found 3 known high or critical vulnerabilities in the packages a fresh install brings (OSV.dev, 3 Nov 2026). Weigh this before you install it.
crates:capnp 0.20.6Cap'n Proto has Undefined Behavior in constant::Reader and StructSchema (GHSA-5w5r-mf82-595p, critical, fixed in 0.24.0)crates:hickory-resolver 0.24.4hickory-resolver: Resolver::lookup() and Resolver::lookup_ip() APIs obscure DNSSEC validation failures (GHSA-5j98-2g5x-46v6, high, fixed in 0.26.2)
See 1 more
crates:rustls-webpki 0.101.7— GHSA-82j2-j2ch-gfr8 (high), fixed in 0.103.13
You need first
- A GPU is required
What it does
Good for
- serve streaming and non-streaming completions over OpenAI v1 API
- generate a 2048x2048 image
- transcribe audio via POST /v1/audio/transcriptions
An alternative to vLLM, llama.cpp and NInfer.