Inkling Mini Hits 300-350 tok/s Locally on RTX 5090
Benchmarking by user willcb shows vLLM + DSPark with FP8 KV cache runs Thinking Machines' Inkling 2 Mini at 300-350 tokens per second on an RTX 5090.
2026-11-22 ~ 2026-11-22 · 2 related posts
- Inkling Mini on a 5090: vLLM+DSPark+FP8 KV hits 300-350 tok/s — willcb · 2026-11-22
1 near-duplicate retellings: willcb