Arc: a ~70 MB streaming speech model that transcribes links names and metadata to your own list as they're spoken, with no second model
Arc-Transcriber is a ~70 MB streaming speech model that performs ASR-NL in one step, transcribing speech every 100 ms. It identifies linked names, numbers with values, speakers, roles, and sound events, flagging new names not on a provided list. The model also detects if speech is addressed to an assistant. Used on-device in a Mac app, it can prime voice search over 61k books with matching titles while the user is still speaking.
Time & source
Times shown in UTC
Display time zone: UTC
Local time zone unavailable; showing UTC.
PublishedOffset at this time: UTC+0Oct 29, 2026, 22:55 UTC
IngestedOffset at this time: UTC+0Oct 30, 2026, 00:00 UTC
- Published
- Oct 29, 2026, 22:55
- Ingested
- Oct 30, 2026, 00:00
- Source type
- Dev community
- Tier
- Community
- Source status
- Sync delayed
Tier is a per-source editorial setting, not a per-item score.
I built Arc-Transcriber, which does ASR-NL in 1-step. Every 100 ms, one model returns the transcript plus:
- Names linked to a list you give it (a roster, contacts, product names); new names are flagged "not on the list" - Numbers with their values (scores, clocks, money, dates) - Speakers and roles; - Sound and channel events (telephone audio, clipping, poor audio) - Whether the speech was addressed to an assistant;
The 200+ labels are trained heads on the shared encoder states: queries per frame, span, utterance, speaker and event, emitted every 100 ms.
The usual ASR, then NER, then entity-linking and then LLM chain is collapsed into the model's own heads, so live "fact-checking" against your list happens inside the stream, not in a second pass.
Currently it has two variants Arc mini (which uses moonshot pretrained encoder) and second Arc-v1 (which uses parakeet pretrained encoder).
Weights: huggingface.co/WhissleAI/arc-v1.1 Code and CLI: github.com/WhissleAI/arc_transcriber
We use it on-device in our Mac app. Voice search over 61k books primes Arc with matching titles while you're still talking. Here is a video to give a glimpse on how it works.