Coqui STT
The community-carried successor to Mozilla DeepSpeech: trainable, streamable, small-footprint STT — now unmaintained.
Retired
Coqui STT appears to be discontinued or shut down. This page is kept as an archive record and is excluded from search engines.
About
Coqui STT was the speech-to-text toolkit built by the ex-Mozilla machine-learning team that spun out as Coqui in 2021, evolving the DeepSpeech codebase with a v1.0 release that year. It offered multi-GPU training, streaming inference, small-footprint acoustic models and bindings for several languages. Its own README now states the project is no longer actively maintained and the hosted Model Zoo has been shut down, with attention shifted toward Whisper-era models; Coqui the company shut down in December 2023. The code stays public under MPL-2.0 (2,604 stars) and pretrained models remain in the coqui-ai/STT-models releases.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
No independent benchmarks found; historical results live in the repo documentation.
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 5/10
Scored 5/10 (verified tier, metrics-v1). Evidence: coqui-ai/STT README verified stating 'no longer actively maintained' with Model Zoo hosting stopped; shutdown confirmed in the maintainer discussion; MPL-2.0 via GitHub API. Strengths: fully open, streaming-capable, small models, training pipeline intact. Docked for: (1) unmaintained since ~2021 with the company gone; (2) no hosted option and no current accuracy story. License: MPL-2.0.
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 5.0/10
- Maintenance · recency, releases, issue load
- 1.0/10
- Openness · license + docs
- 9.0/10
- Manual review · sanity read + red-flag check
- 6.0/10
- · README states 'no longer actively maintained'; Model Zoo hosting stopped
- · Coqui the company shut down December 2023
- · 2,604 stars; v1.0 released 2021
- · Pretrained models remain in coqui-ai/STT-models releases