About
UzCaption.
UzCaption is a software company. We build AI subtitles for the Uzbek language — a language that general-purpose transcription tools consistently get wrong. The product is sold online as a five-tier monthly subscription billed by the minute.
What we do
UzCaption converts Uzbek speech into accurate, correctly timed subtitles. A user sends a video and receives either a video with captions burned in or a standalone SRT file that opens in Premiere Pro, DaVinci Resolve, CapCut or YouTube. Subtitles can also be translated into other languages.
The recognition is trained and tuned specifically for Uzbek — including regional dialects, and the written details that matter, such as the apostrophes in o' and g', punctuation, and where a sentence should break across two caption lines. This is the part that off-the-shelf multilingual tools do not solve.
How the product is delivered
The Android app is on Google Play; the iOS app is coming soon. Every channel runs on the same account, the same plan and the same speech engine. Video language: Uzbek, Russian, English, Tajik and Karakalpak; translation into 6 languages.
How the business works
UzCaption is a digital-native subscription business. There is no physical product, no shipping and no offline sales channel: a customer signs up online, pays online, and the service is delivered as software over the internet.
| Model | Self-serve SaaS — five-tier monthly plan (1/3/6 months), billed by the minute |
|---|---|
| Free plan | 5 minutes every month, no card required |
| Paid plans | Mini — 39,000 UZS/mo ≈ $3 (20 min/mo) · Pro — 89,000 UZS/mo ≈ $7 (100 min/mo, most popular) · Max 5× — 149,000 UZS/mo ≈ $12 (500 min/mo) · Max 20× — 399,000 UZS/mo ≈ $32 (2,000 min/mo) |
| Customers | Video creators, editors, journalists, teachers and media teams working in Uzbek |
| Infrastructure | Cloud-hosted. Speech recognition, rendering and delivery are fully automated — no manual step between payment and result. |
Full pricing is public on the product page, and the legal terms are in our public offer and privacy policy.
The team





Backed by
- UzCombinator
- NVIDIA Inception
- AWS Activate
- Cloudflare for Startups
- PostHog for Startups
- GitHub for Startups
- Higgsfield
- Yoshlar Ventures
Why Uzbek
Uzbek is spoken by roughly 35 million people, yet it sits far down the priority list for global speech-recognition providers. Accuracy on Uzbek audio is poor, dialects break it, and the Latin-script orthography is handled inconsistently. That gap is the entire reason UzCaption exists.
Beyond the product, we are building an Uzbek speech corpus so that recognition keeps improving — with explicit user consent, and with a public correction channel so customers can report mistakes directly.
Company and contact
| Trading name | UzCaption (UzCaption AI) |
|---|---|
| Website | caption.uz — open from anywhere, pay with Uzbek cards |
| Support | @jilo_support |
| Product support | Chat in Ijodxona |
| Languages | Uzbek, Russian, English |