All tasks / Models
omniASR-LLM-7B-v2
Meta's open speech recognition model suite covering more than 1,600 languages.
- Meta (FAIR)
- Open weights
Where it ranks
- BTranscriptionModels · Local · 8.1% on Hugging Face Open ASR Leaderboard (English, short-form average WER) (secondary benchmark)
Facts
- Vendor
- Meta (FAIR)
- Family
- Omnilingual ASR
- Released
- —
- Weights
- Open: download and run it yourself
- Parameters
- 7B
- GPU memory to run locally
- ~16 GB (Derived: about 7.00B parameters is roughly 14.0 GB of bf16 weights plus runtime overhead; 7B taken from the model name; not an official requirement.)
- Access
- Free (open weights)
- Last verified
- 19 Oct 2026
Benchmark results
| Benchmark | Result | Entry on the leaderboard | As of | Source |
|---|---|---|---|---|
| Hugging Face Open ASR Leaderboard (English, short-form average WER) | 8.1% | facebook/omniASR-LLM-7B-v2Average WER over the board's default datasets (English short-form, version 17-10-2026); rank 45 of 56; read from the Space's /config table data. | 17 Oct 2026 | huggingface.co |
Where you can use it
No tracked tool lists this model yet; as open weights it runs in local runtimes.