Illustration, AI-generated from the story text. Tap a panel to read the story through it.

Salesforce published SOV-Transcribe-1 under on , a model with about 4 billion active parameters out of 26 billion that transcribes seven languages and takes plain-language instructions on formatting and vocabulary.

It scores a 2.28% on Artificial Analysis, first among open-weight non-streaming models and seventh overall, ahead of Nvidia's Nemotron 3.5 Transcribe at 2.81% and Mistral's Voxtral Small at 2.77%, and 8.21% on call-center audio across 14 English accents.

The parameter counts differ by source (26B or 26.6B total, 4B or 4.6B active).

A contact center that could not send audio to a vendor's cloud can now run a leading transcriber on its own hardware.
Salesforce releases an open-weight speech recognition model that tops the open field on Artificial Analysis
Salesforce published SOV-Transcribe-1 under Apache 2.0 on Hugging Face, a mixture-of-experts model with about 4 billion out of 26 billion that transcribes seven languages and takes plain-language instructions on formatting and vocabulary. It scores a 2.28% word error rate on Artificial Analysis, first among open-weight non-streaming models and seventh overall, ahead of Nvidia's Nemotron 3.5 Transcribe at 2.81% and Mistral's Voxtral Small at 2.77%, and 8.21% on call-center audio across 14 English accents. The parameter counts differ by source (26B or 26.6B total, 4B or 4.6B active). A contact center that could not send audio to a vendor's cloud can now run a leading transcriber on its own hardware. SourcesAC
Sources
- AHugging Face: SOV-Transcribe-1huggingface.co
- CCrypto Briefing: Salesforce SOV-Transcribe-1cryptobriefing.com
This story ran in the Morning Brief of November 14th 2026.