This product was not featured by Product Hunt yet. It will not be visible on their landing page and won't be ranked (cannot win product of the day regardless of upvotes).
Most speech recognition models are trained on clean English and fall apart on Arabic — especially dialects and the code-switching people actually use every day. Audar is built Arabic-first: open-weight ASR models covering MSA, dialectal Arabic, and code-switching, trained on real-world audio instead of studio recordings. Inspect the weights, fine-tune on your own data, and deploy without lock-in. Built for developers and researchers building for the 400M+ people who speak Arabic.
Hi Product Hunt 👋
I'm Sia from the Audar team.
If you've ever tried to run Arabic through an off-the-shelf speech model, you know the feeling: it does okay on formal, textbook Arabic, then completely falls apart the moment someone speaks in a dialect or switches between Arabic and English mid-sentence — which is how most people actually talk.
That gap is why we built Audar. Arabic is spoken by hundreds of millions of people, and it deserves speech recognition built for how it's really used, not a clean-data afterthought. So we went Arabic-first: open-weight ASR models covering MSA, dialectal Arabic, and code-switching, trained on messy real-world audio.
Everything is open-weight on purpose. Inspect the models, fine-tune them on your own domain, deploy them however you want — no API lock-in.
We'd genuinely love your feedback, especially if you work with Arabic audio or low-resource languages more broadly. What breaks? What would you want it to handle next? Drop your toughest audio at us — dialects, noise, whatever — and let's see how it holds up.
Happy to answer anything today. 🙏
Finally an ASR that doesn't choke on my Levantine Arabic with English mixed in, and being able to peek at the weights locally is a real plus for debugging.
Adding a built-in streaming inference mode with WebSocket support would be a nice next step so we can pipe microphone audio in directly without managing chunking on our end.
Real useful work, finally something that doesn't choke on Moroccan darija mixed with French. One thing that would help a lot for our team: ship a small streaming WebSocket demo with partial transcripts and a confidence score, so we can see latency on noisy mobile audio before committing to a fine-tune.
Audar-ASR-V1 was submitted on Product Hunt and earned 0 upvotes and 6 comments, placing #115 on the daily leaderboard. Most speech recognition models are trained on clean English and fall apart on Arabic — especially dialects and the code-switching people actually use every day. Audar is built Arabic-first: open-weight ASR models covering MSA, dialectal Arabic, and code-switching, trained on real-world audio instead of studio recordings. Inspect the weights, fine-tune on your own data, and deploy without lock-in. Built for developers and researchers building for the 400M+ people who speak Arabic.
Audar-ASR-V1 was featured in Artificial Intelligence (473.8k followers), GitHub (41.3k followers) and Audio (2.1k followers) on Product Hunt. Together, these topics include over 136k products, making this a competitive space to launch in.
Who hunted Audar-ASR-V1?
Audar-ASR-V1 was hunted by Sia He. A “hunter” on Product Hunt is the community member who submits a product to the platform — uploading the images, the link, and tagging the makers behind it. Hunters typically write the first comment explaining why a product is worth attention, and their followers are notified the moment they post. Around 79% of featured launches on Product Hunt are self-hunted by their makers, but a well-known hunter still acts as a signal of quality to the rest of the community. See the full all-time top hunters leaderboard to discover who is shaping the Product Hunt ecosystem.
Want to see how Audar-ASR-V1 stacked up against nearby launches in real time? Check out the live launch dashboard for upvote speed charts, proximity comparisons, and more analytics.
Sia He