SOFTWAREvLLM Project

vLLM

Easy, fast, and cost-efficient LLM serving for everyone.

Visit site View stories

Open-source library and inference/serving engine for LLMs, focused on high throughput and memory efficiency.

Recent stories

1 linked story

releaseSECONDARY2026-03-26

Cohere launches Transcribe 03-2026 with 14 languages and Apache 2.0 weights

Cohere released a 2B speech-to-text model with 14 languages and top Open ASR scores, and upstreamed encoder-decoder optimizations to vLLM in the same launch. It is a self-hosted ASR option, so test accuracy and throughput on your own speech workload.