Blog | Smean ស្មៀន

Smean Creative Challenge - register for an extra 15 minutes free

Mon, 08 Jun 2026 00:00:00 +0000

We’re launching the Smean Creative Challenge - a chance to show what you can build, record, or create with Khmer speech AI.

Register below and we’ll add 15 minutes of free transcription credit to your account automatically. No card required, no strings attached.

Register for the challenge

Loading…

If the form doesn’t load, open it in a new tab.

Once you submit, your extra 15 minutes will appear in your Smean account within 24 hours. Questions? Reach us at info@smean.ai.

5 tips for getting better transcription results

Tue, 26 May 2026 00:00:00 +0000

We get asked a lot about how to squeeze more accuracy out of SMEAN - usually right before someone uploads a two-hour meeting recorded on a phone left face-down on a table. The honest answer is that what happens before the file reaches us matters more than anything we do on the model side. A few small habits during recording move accuracy further than most model upgrades.

1. Record at 16kHz mono or higher

Phone defaults are fine for most situations. The trouble starts with heavy compression - anything below 64kbps and the model starts losing the high-frequency detail it needs to tell similar Khmer phonemes apart. If your recorder offers WAV or FLAC, prefer it over MP3 for anything you actually care about. The file is bigger; the transcript is better.

How AI is transforming Khmer language documentation

Tue, 26 May 2026 00:00:00 +0000

For most of the 20th century, Khmer lived more in the air than on the page. Court proceedings, hospital intake, university lectures, NGO field interviews, family histories - spoken in full, captured in fragments. The cost of writing it all down was simply too high.

At SMEAN, we’ve watched that cost collapse. An hour of Khmer audio that used to take a human transcriber a full working day - and somewhere around $80 to produce - now runs through our pipeline in minutes for less than the price of a coffee. That shift sounds technical, but it reshapes who gets to keep a record of what they said.

Introducing SMEAN 2.0: a new workspace, a consistent design, and full cross-device support

Tue, 26 May 2026 00:00:00 +0000

Today we’re opening the beta for SMEAN 2.0.

This is the biggest release we’ve shipped since launching in early 2025. It isn’t a model update - the transcription engine you already know is still doing the heavy lifting underneath. What’s new is everything around it. The workspace, the visual identity, and the way SMEAN behaves across your devices have all been rebuilt to live in one consistent standard.

What’s new in the beta

A redesigned workspace

For a long time, SMEAN was a list of transcripts. That worked when most people had ten files. It stops working at a hundred, and it stops working entirely when a team is involved. The 2.0 Workspace is built as a real working surface:

Smean 2.0 is in beta - new design, faster transcription, and a redesigned workspace

Tue, 26 May 2026 00:00:00 +0000

Today we’re opening Smean 2.0 beta to early users.

It’s not just a model update. The whole product has been redrawn around what we’ve learned from a year of running the service in production - what people actually do with transcripts, where the old UI got in the way, and what was missing for teams larger than two or three people.

What’s new

A new design system

The 2.0 surface is built on a refreshed design system: editorial type pairings, a 4px grid, semantic color tokens that swap cleanly across light/dark and across our product slots (Scribe, Studio). Every surface - onboarding, the workspace, the marketing site - speaks the same visual language now.