SOV-Transcribe-1
SOV-Transcribe-1 is available now for vision with open-source; evaluate it while provider pricing coverage matures.
- Family
- SOV-Transcribe
- Released
- 2026-11-12
- Parameters
- 26.6B (4.6B active)
- Architecture
- Mixture of Experts
- Specialization
- speech-recognition
- Openness
- Open source
- License
- Apache 2.0OSI-approvedPermitted
- Weights
- Available
- Code
- Unknown
- Training
- Fine-tuned
Ethical AI solutions for customer empowerment
No tracked provider pricing is available yet.
Ethical AI solutions for customer empowerment
No tracked provider pricing is available yet.
About
SOV-Transcribe-1 is Salesforce's first open-weight speech model, announced November 12, 2026 by its Sovereign AI team: a speech-augmented language model for non-streaming automatic speech recognition. It pairs a Whisper large-v3 audio encoder and an audio projector with the Gemma 4 26B A4B mixture-of-experts language model, for 26.6B total and 4.6B active parameters. It transcribes English, French, German, Spanish, Italian, Portuguese and Dutch from 16 kHz audio of up to 30 minutes (tested), and takes optional natural-language instructions for verbatim or written-form style and context biasing with names or domain vocabulary.
Providers
No active provider routes are tracked for this model yet.
Capabilities
Benchmark peer barsfor Vision
No task-mapped benchmark peers are available for this model yet.
Migration checks
No linked migration route is available for this model yet.
API versions
salesforce-sovereign-intelligence/SOV-Transcribe-1