Hugging Face and Cerebras put Gemma 4 31B into an open-source voice AI stack
_akhaliq · x · 2026-07-21
Hugging Face and Cerebras say developers can now use Gemma 4 31B as the core model in a fully open-source speech-to-speech stack.
- The pitch is ultra-fast inference for voice AI, aimed at powering existing voice apps.
- The stack is cascaded and open source, so builders can plug it into their own systems rather than wait on a closed vendor pipeline.
More from Infra
- llama.cpp lands Flash Attention tuning for RDNA4, big prefill gains on AMD — pmttyji · 2026-09-11
- Your p99 latency benchmark may be lying: a deep dive into coordinated omission — Franc0Fernand0 · 2026-09-11
- Running MiniMax H3 on 12GB VRAM: quantization, Turbo LoRAs and attention backends compared — Possible_Mood676 · 2026-09-11
- Spomin: live KV cache compaction squeezes 500k tokens of context into 180k resident — wgaca2 · 2026-09-11
- PiPNN nearest-neighbor search wins three awards, up to 78x faster index building — khademinori · 2026-09-11
- M.2-Oculink eGPU Link Silently Downgrades to PCIe Gen1 — Here's How to Check — El_90 · 2026-09-11