Egyptian and Nigerian English Added to Hugging Face ASR Leaderboard
Egyptian and Nigerian English have joined Hugging Face’s Open ASR Leaderboard, expanding the English varieties represented on the speech recognition evaluation platform. The leaderboard lets users explore and compare model performance across datasets.
- Egyptian and Nigerian English joined Hugging Face’s Open ASR Leaderboard.
- The leaderboard compares English speech recognition model performance.
- It presents speech model performance across datasets.
- The source gives no scores, rankings or model-specific results for the update.
Two English varieties added
Hugging Face’s announcement is titled “Egyptian and Nigerian English join the Open ASR Leaderboard.” It identifies the two varieties as the latest additions to the leaderboard, which is presented as a place to explore and compare English speech recognition model performance.
The source does not provide scores, rankings, model names or details of the evaluation datasets associated with the new entries. It also does not state whether the additions change the position of any model on the leaderboard.
A broader view of speech recognition performance
The Open ASR Leaderboard describes its focus as speech model performance across datasets. Adding Egyptian and Nigerian English gives the platform two more varieties through which users can examine automatic speech recognition results, according to the announcement.
That focus makes the update relevant to developers and researchers comparing speech recognition systems. Results shown for one variety of English do not by themselves describe performance for every other variety, so the presence of additional evaluation categories can help users inspect performance in a wider set of speech contexts. The source, however, does not make a broader accuracy claim about the models.
Part of Hugging Face’s speech evaluation work
The announcement appears alongside Hugging Face’s other speech-related leaderboard and blog material. The page lists the Open TTS Leaderboard, described as covering scalable evaluation for multilingual text-to-speech and voice cloning, and an August 28, 2026 article titled “The Open ASR Leaderboard Adds Its First Global South Language.”
The page also links to RW-Voice EQ ASR, described as a place to explore and compare English speech recognition model performance. These related entries place the Egyptian and Nigerian English update within Hugging Face’s wider collection of tools and materials for evaluating audio and speech models.
What the update means for users
For people using the Open ASR Leaderboard, the immediate change is the availability of Egyptian and Nigerian English as part of its English speech recognition coverage. Users can use the leaderboard to explore model performance across datasets, while the announcement provides no further information about access requirements, release timing beyond the page date, or the underlying data.
Hugging Face: Egyptian and Nigerian English join the Open ASR Leaderboard
