Methodology
Facts
Every field (license, commercial use, parameters, memory, streaming, cloning, languages) comes from the model's Hugging Face card, its GitHub README, or its LICENSE file, and links to that source. If the source doesn't say, we show Unknown instead of guessing. Release date is the Hugging Face repo creation date unless noted.
Commercial use is Yes when the license permits it (attribution may still be required, e.g. CC-BY-4.0), No for non-commercial licenses, and Conditional when it depends on revenue caps, copyleft (GPL) terms, or contacting the vendor. This is not legal advice. Read the license.
Benchmarks
- RTF (real-time factor) = wall-clock processing time ÷ audio duration. 0.05 means 20× faster than real time.
- TTS: time to first audio chunk, and RTF to synthesize a fixed ~10 s English passage.
- STT: latency to the first transcribed segment and RTF on a fixed English clip, plus WER against the reference text.
- Memory: peak resident memory of the process (RSS); peak GPU memory when run on GPU.
- One warm-up run, then the median of 3 runs. Hardware is labeled on every result.
The harness is open source (bench/run.py) so anyone can reproduce a number.