Salesforce SOV-Transcribe-1 Speech Transcription Model Weights Released

Posted Nov 13, 07:02 As of Nov 12, 2026

The Hugging Face model card includes instructions for running the model with Transformers and vLLM, along with DFlash speculative decoding draft settings. Benchmarks and parameter counts were not confirmed in the official page body.

What happened

Based on the Hugging Face model card (official from Salesforce)
Transformers serving command and vLLM serving command are listed.
DFlash speculative decoding draft settings are included.
The weights have been published on Hugging Face.
Benchmark rankings, parameter structure, and license mentioned only in secondary reports were not used as official figures.

Our take

Users who want to experiment with local speech recognition on a PC or Mac can check the execution settings in the model card directly. However, memory requirements, commercial use conditions, and actual accuracy are not confirmed on the official page and remain unknown.

Sources and related posts

Salesforce (Hugging Face 모델 카드) Primary source

llama.cpp b12322 Pre-release, Including Graph Memory, SYCL and CUDA Fixes (github.com)

The b12322 pre-release was published on November 12 at 18:59 UTC, and consecutive builds from b12308 through b12322 were posted the same day. The included changes are the fixes listed on the release page.

Local AI| Nov 13, 07:02 |1 sources |Our take included

llama.cpp b12241 fixes FastRPC handle truncation on Hexagon (github.com)

The ggml-org/llama.cpp b12241 release is a minor update that fixes a bug where 64-bit FastRPC handles were truncated on close in Hexagon. The release page does not list new model support or user-facing features.

Local AI| Nov 12, 07:02 |1 sources |Our take included

Ollama v0.40.16 Stable Release Notes Memory Savings on Apple Silicon (github.com)

Ollama has released the v0.40.16 stable tag. The official release notes state that memory usage for Qwen3.8 Flash Next on Apple silicon has decreased, and they include CUDA 13.5 support for Windows on ARM.

Local AI| Nov 11, 07:02 |1 sources |Our take included