Developers Push Inkling 2 Mini to ~200 tok/s on Dual 3090s
Developers ported Thinking Machines' newly open-sourced Inkling 2 Mini to vLLM, boosting decoding speed to about 200 tok/s on dual RTX 3090s, up from 50 tok/s out of the box, while a quantized version runs at nearly 60 tok/s on a single 4090.
2026-11-21 ~ 2026-11-22 · 2 related posts
- Inkling 2 Mini quantized to 13.9GB runs at ~60 t/s decode on a single RTX 4090 — indrasmirror · 2026-11-21
- Dev ports Inkling 2 Mini to vLLM, hitting ~200 tok/s decode on 2x RTX 3090 vs 50 on llama.cpp — peaster_ · 2026-11-22