Model serving Inference Optimization Open Models Inkling

TokenSpeed

Speed-of-light LLM inference

TokenSpeed is an open-source, production-oriented LLM inference engine for low-latency OpenAI-compatible serving of agentic workloads, with a compiler-backed modeling layer, scheduler, KV-cache resource management, and pluggable kernel system including optimized MLA kernels.

Recent stories

1 linked story

newsSECONDARY2026-07-16

Inkling adds early llama.cpp serving via 1-bit GGUF

Inkling's 1-bit GGUF ran in llama.cpp at 30–40 TPS, and TokenSpeed added day-zero support with a flat KV cache pool. Arena posts put Inkling #10 among open models in frontend code and text, while docs drew scrutiny.