Some speech recognition must happen on the device, on the go. Audio8 ASR 0.1B brings practical automatic speech recognition to phones, wearables, recorders and other constrained hardware.
A small model is not just a shrunken one, but also requires zero network, keeps sensitive audio on the device, and still delivers useful recognition.
| Item | Details |
|---|
| Model | Audio8 ASR 0.1B |
| Scale | 0.1B parameters |
| Deployment focus | Private local inference |
| Runtime | CPU / edge / mobile / Apple ANE |
| Languages | 7 |
| Best suited for | Voice input, field notes, offline captions, privacy-sensitive apps |
Satisfactory recognition on the go.
When a model lives on the device, parameter count, peak memory and power draw determine whether the product is viable at all. Audio8 ASR 0.1B is designed for local real-time transcription, stable responses and sustained low-power operation.
| Capability | Benefit |
|---|
| Portable ONNX Runtime form | No CUDA or PyTorch dependency; straightforward to deploy on a device |
| INT8 / INT4 quantization | Cuts storage and memory so lower-end devices can run it |
| Apple ANE path | Runs recognition on the Neural Engine for better battery life |
| Fully local processing | Audio never leaves the device |
Use cases and getting started
Use it for mobile voice input, interview and meeting notes, offline captions and privacy-sensitive apps. To start, choose the ONNX Runtime or Apple ANE build, connect the Swift or C interface, and transcribe locally.
The family at a glance
Choose Audio8 ASR 3B for the highest quality on complex audio. Consider Audio8 ASR 0.6B when GPU resources and cost need to be balanced.
- Audio8-ASR-3B · High-accuracy flagship: complex audio, multilingual, high-quality transcription
- Audio8-ASR-0.6B · Balanced choice: quality, speed, and cost
- Audio8-ASR-0.1B · Lightweight edge model: local recognition on phones and edge devices
Limitations and responsible use
Automatic transcription can be affected by noise, accents, overlapping speech, specialist terms, and audio quality, and should not be treated as an unreviewed factual record. Medical, legal, financial, and other high-risk uses require human review, and the deployer is responsible for the final output.
Obtain authorization before processing personal voices, conversations, or sensitive information and comply with applicable laws such as GDPR. Audio and text should be encrypted, access-controlled, and retained only as long as necessary.
Audio8 ASR 0.1B
Useful, private recognition within a truly on-device budget.
