# Rumble-2-Flash

> Rumble-2-Flash is Reson8's 0.5B open-weight speech-recognition model for 45 languages. It takes audio and returns text, with a 65,536-token text context. Listed as upcoming.

**ID**: `reson8/rumble-2-flash`  
**Creator**: reson8  
**Category**: text  
**Context**: 66K tokens  
**Status**: Coming soon — announced, not yet routable on AnyRouter; requests return 404 `model_not_yet_available`.  
**Note**: Coming soon. Reson8 published Apache-2.0 weights on 2026-11-18; no inference API serves them yet. We will list prices and route requests once one does.  
**Pricing**: Not published yet  
**Released**: 2026-11-18  
**Links**: [Weights on Hugging Face](https://huggingface.co/Reson8/Rumble-2-Flash) · [Announcement](https://x.com/raoulritterr/status/2123053360773271563)  
**Web page**: https://anyrouter.dev/model/reson8/rumble-2-flash

**Input modalities**: audio  
**Output modalities**: text  

**Capabilities**: transcription, audio-input
