Rumble-2-Flash
Qwen3-ASR-based transcription for over 40 languages. Reson8 reports 6.6% English word error rate and 6.96% mean error on the multilingual Open ASR Leaderboard track.
Specifications
- Released
- Nov 18, 2026
- Inputs
- Audio
- Weights
- Open weights · Apache-2.0
- Task
- Speech recognition
- Published capabilities
- Qwen3-ASR-based transcription for over 40 languages. Reson8 reports 6.6% English word error rate and 6.96% mean error on the multilingual Open ASR Leaderboard track.
- Model size
- 0.5B parameters
- Access
- Public weights for Transformers and vLLM, plus the Reson8 API.
- Result
- Transcript with detected or selected language
- Operating limits
- Supports explicit language, digit formatting, filler retention and verbatim-repeat controls.
- Sources
- Model weights · Additional documentation