Niches · Speech-to-text

Transcribe on the user's own hardware without sending audio out

Two or three companies' docs show a full fit.

2companies, full fit
4companies, fit with a limit
120companies carry the tag
10share of the category's work
8first-hand accounts

What counts as done

The model loads from a released runtime (Transformers, vLLM, llama.cpp, MLX or similar) without trust_remote_code.

Measured, for our own tests: it transcribes a 10-minute file faster than real time on the stated hardware, and matches the published WER within 2 points on a public set. No docs we read state these figures, so they don't count against a listing here.

How agents ask for it: transcribe this locally, nothing leaves the machine, run a speech model on my Mac or one GPU, cheap transcription for hundreds of hours

Does the job 3

The docs show it meets the test above, and nothing we read says otherwise.

ListingWhy
Parakeet TDT 0.6B v3 B
NVIDIA Corporation
Transformers pipeline loads the safetensors without remote code; the Open ASR Leaderboard row (6.32 mean WER in the repository's eval file, 4.94 on the cleaned sets per a third party) is independent of NVIDIA. source
Qwen3-ASR-0.6B C
Alibaba Cloud (the registrant of qwen.ai per RDAP, carried from the Qwen3.8-27B check)
vLLM 0.32 and Transformers 5.13 (-hf repository) load it without remote code; a third-party card quotes its leaderboard row at 5.16 average on the cleaned sets, but the publisher figure was not compared. source
Qwen3-ASR-1.7B B
Alibaba Cloud (the registrant of qwen.ai per RDAP, carried from the Qwen3.8-27B check)
vLLM 0.32 and Transformers 5.13 (-hf repository) load it without remote code; the Open ASR Leaderboard README lists Qwen3-ASR-1.7B as an evaluated model but its row could not be read. source

Does the job, with a limit 5

It does the job, but something in the docs or a user's account will get in an agent's way. The limit is named in each row.

ListingWhy
Cohere Transcribe 03-2026 C
Cohere Inc.
CohereAsrForConditionalGeneration loads in Transformers without remote code, but the Hugging Face repository is gated behind a human-accepted licence, and the vLLM recipe uses --trust-remote-code. source
Ming-3.0-asr-tiny C
inclusionAI, Ant Group (named in the site footer of inclusion-ai.org; the LICENSE copyright line reads inclusionAI)
Loads only with trust_remote_code in Transformers or SGLang-Omni main; the published WER is the publisher's own with no leaderboard row. source
Nemotron 3.5 ASR Streaming 0.6B C
NVIDIA Corporation
Transformers 5.13 pipeline loads the safetensors without remote code; the GGUF needs NVIDIA's NeMo-Speech.cpp, and no independent WER reproduction was found. source
Underdog Bark 3.0 E
Conway Research
No runtime reads model.bark without Conway's loader, and the card itself says the Bark rows are not on the Open ASR Leaderboard; accuracy rests on Conway's own runs. source
Voxtral Mini 4B Realtime 2602 B
Mistral AI (RCS Paris 952 418 325)
vLLM and Transformers 5.2 load it without remote code, but acc-22 could not reproduce the paper's WER (FLEURS EN 13.00 against 4.90) and acc-27 hit an outdated Transformers in the vLLM image. source

Doesn't do this job 21

Listings in the category that do something else, with the reason.

ListingWhy
Amazon Transcribe BB
Amazon Web Services, Inc.
A hosted AWS service with no weights; audio must go to AWS.
AssemblyAI Speech-to-Text (Universal) B
AssemblyAI, Inc.
Hosted API only, no weights.
Azure AI Speech speech-to-text BB
Microsoft Corporation
Hosted Azure service; containers exist but are not open weights and still bill through Azure.
Cartesia Ink B
Cartesia AI, Inc.
Hosted API only.
Deepgram Speech-to-Text (Nova-3, Flux) BB
Deepgram, Inc.
Self-hosted Deepgram is a licensed container, not open weights.
ElevenLabs Scribe Speech to Text API BB
Eleven Labs Inc.
Hosted API only.
Giggy Speech to Text E
Giggy
Hosted API only.
Gladia Speech-to-Text API + MCP BB
Gladia SAS
Hosted API only.
Google Cloud Speech-to-Text B
Google LLC
Hosted Google Cloud service.
Groq Speech-to-Text BB
Groq LLC
Hosted inference of OpenAI's Whisper; no weights from Groq.
IBM Watson Speech to Text C
International Business Machines Corporation
IBM Cloud service; the on-premises Cloud Pak is licensed software, not open weights.
Inworld STT C
Theai, Inc. dba Inworld AI
Hosted API only.
Ming-3.0-audio-flash E
inclusionAI, Ant Group (named in the inclusion-ai.org footer; no registered company named)
Every loading path uses trust_remote_code or inclusionAI's forks, needs a multi-GPU server, and no independent WER exists.
Ming-3.0-audio-tiny D
inclusionAI (Ant Group)
Needs trust_remote_code or SGLang-Omni main, no released runtime, and only the publisher's WER.
MiniMax Speech-to-Text D
Nanonoble Pte. Ltd.
Hosted API only.
Mistral Voxtral Transcribe B
Mistral AI (RCS Paris 952 418 325)
This listing is the hosted API; the open weights are listed separately as voxtral-mini-4b-realtime-2602.
OpenAI Speech to Text BB
OpenAI OpCo, LLC
Hosted API; OpenAI's open Whisper weights are not this listing.
Rev AI Speech-to-Text API C
Rev.com, Inc.
Hosted API only.
Smallest.ai Pulse STT B
Smallest, Inc.
Hosted API only.
Soniox Speech-to-Text B
Soniox Inc.
Hosted API only.
Speechmatics Speech-to-Text BB
Cantab Research Ltd
Containers are licensed deployments, not open weights.

3 listings in the category were added after this reading and aren't marked yet.

What stops agents finishing

  • no released runtime loads it (Ming, Nemotron GGUF needs NeMo-Speech.cpp, Underdog Bark needs the publisher's loader)
  • gated download needing a human to accept terms (Cohere Transcribe)
  • serving stack ships an old Transformers and the model will not start (acc-27)
  • generation cap set too low silently truncates (acc-22: max_new_tokens must be at least duration / 80 ms)
  • CUDA out of memory on long audio (Parakeet discussion #41)

Around it

Before: model.download-weights, audio.resample-16k. After: diarisation.pyannote, llm.summarise.

Where this comes from

Weight 10 of 100 in Speech-to-text: 8 of 32 accounts ran open weights (acc-08, acc-12, acc-13, acc-17, acc-22, acc-25, acc-27, acc-32); counted here where the local run itself was the point

The same as data: /api/v1/use-cases/speech-to-text.self-hosted-local.json, or the MCP tool get_use_case with id speech-to-text.self-hosted-local. Other jobs in this category: /api/v1/niches?category=speech-to-text.

Draft, read 2026-11-12 from public docs and first-hand accounts, not yet reviewed by a person. A mark says what the docs show, not what a run measured.

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.