Sovereign cloud
Servers, storage and networks on Kawkav’s own fleet — not a reseller’s.
Private AI
Run open models on sovereign GPUs — nothing leaves your cloud.
Own your comms
A phone system, eSIM data and video rooms — under your own name.
Run the company
One sovereign platform for the back office, the shop, the farm and the plant.
The mission
Media tools, training and the sovereignty thesis behind Kawkav.
Media · AI virtual producer
A complete AI studio for creators and newsrooms. Give it a script and it speaks — in 40+ languages including native Arabic — puts a presenter on screen, renders broadcast-ready video and podcasts on our GPUs, and captions every word. Produce at the scale of a channel with the team of a desk.
Write or paste a script and the studio handles the rest — voice, presenter, render and captions — then hands you a file ready to publish. Everything runs on Kawkav's own sovereign GPUs, so your footage and scripts never leave your tenant.
Kawkav XTTS turns any script into natural narration across 40+ languages — with genuinely native, well-pronounced Arabic that most engines get wrong. Pick a voice, set the pace, and it reads.
Put a lifelike avatar on screen to deliver the piece to camera, lip-synced to the generated voice. The avatar and video render are in beta today — great for drafts and social, improving fast.
Every render ships with word-accurate captions and a full transcript. Burn them in, export SRT/VTT, or translate the caption track into another language for a second cut.
Voice, avatar, captions and your visuals are composited and rendered on our GPUs into finished MP4. Render is billed by the minute, so you only pay for the footage you actually produce.
Produce one script, publish it in three languages. Translate the copy, re-voice it with a matching XTTS speaker, and re-render — one piece becomes a multilingual release for the whole region.
The same pipeline outputs a finished video, an audio-only podcast episode, or a live-style presenter read for a stream. One studio, three formats, ready to publish anywhere.
Four stages, fully automated on our own GPUs — you write the words, the studio delivers the file.
The teams producing at volume — where a studio, a crew and a translator for every clip simply doesn't scale.
Turn lesson notes into narrated, captioned video courses — in your students' language, without a camera.
Spin up product explainers and social spots in every market you sell to, from one script, on brand.
Convert a wire story into a presenter-read video bulletin within minutes of it breaking — captioned and translated.
Publish daily to YouTube, TikTok and a podcast feed without burning out — voice, edit and captions handled.
The whole studio — XTTS voices, render, transcription — runs on Kawkav's own GPUs under your tenant. Scripts, voices and finished media stay in-house, not shipped to a third-party cloud. Keep your data in-house, publish to the world.
Text-to-speech, captions and transcription are production-ready today. The virtual presenter and full video render are in beta — strong for social and drafts, still improving on long-form and complex scenes. We'll tell you which is which, every time.
Both plans include a monthly allowance of GPU render minutes; extra render is billed by the minute as you use it. Prices exclude VAT.
GPU render is metered in minutes on top of your monthly allowance — you only pay for footage you actually produce. Need a team plan or a high-volume render pipeline? Talk to us →
Genuinely native. Kawkav XTTS was built to pronounce Modern Standard Arabic correctly — including the sounds and stress patterns most Western TTS engines mangle. You can hear it before you commit.
The virtual presenter (avatar) and the full video render pipeline. Text-to-speech, captions and transcription are production-ready. Beta features work well for social clips and drafts and are improving quickly — we label them clearly so you're never surprised.
By the minute of finished output. Each plan includes a monthly allowance of render minutes; once you pass it, further render is metered per minute at a published rate. Audio-only podcasts are far cheaper to produce than full video.
Yes — that's a core use case. Write once, translate the script, re-voice it with a matching XTTS speaker, and re-render. Your caption track translates alongside it, so a single production becomes a multilingual release.
On Kawkav's own sovereign infrastructure under your tenant — not a third-party rendering cloud. Your scripts, voices and finished files stay in-house, and you can export everything at any time.
From €19/month — voice, captions and podcast today; avatar and GPU video in beta. Everything on your own sovereign tenant.