What is TubeTube?
By Arsène Huot, founder of TubeTube · Published · Updated
TubeTube in one paragraph
TubeTube is an AI video studio. You feed it text, either song lyrics or a narration script, and it hands back a finished multi-scene video with no manual editing. One run does the lot: it generates the audio, transcribes it word by word, builds a storyboard, draws a consistent image for each scene, animates them, and assembles the final cut. It was made for narrative content, kids songs, bedtime stories, music videos, where the same character has to stay believable across dozens of scenes.
How does TubeTube work?
A single TubeTube run moves your text through an end-to-end pipeline, picking the best model for each step. The order matters because every later step is timed to the audio:
- Text. You write lyrics (for a music video) or a narration script (for a story or explainer). A narration of about 2,500 characters makes roughly a 3-minute video, and more text just makes more scenes.
- Audio. Suno composes and sings the lyrics, ElevenLabs narrates the script with a voice you choose, or you upload your own mp3. This audio is the spine of the whole video.
- Word-by-word timing. The audio is transcribed and timed word by word, so scene changes can land on the exact phrasing of the song or narration.
- Storyboard. The timed text is broken into one scene every few seconds.
- Consistent visuals. Each scene gets one image, generated sequentially with the previous scenes as context, plus up to 5 pinned reference images and a choice of 113 visual styles, so the same character and world carry through.
- Animation. A video engine animates each scene into motion.
- Assembly. The clips are edited together automatically to match the audio timing, with optional background sounds, into a final video.
This is the same end-to-end process described in our guide to making faceless YouTube videos with AI, and the reason TubeTube reads as a long-form video generator rather than a single-clip tool. A narrated video can also be dubbed into up to 5 languages at once, billed per minute per language, with failed languages refunded automatically; songs aren't dubbed. Every term used here, storyboard, dubbing, video engine, is defined in the AI video glossary.
Who builds TubeTube?
TubeTube is founded by Arsène Huot, a French creator who runs real multilingual faceless YouTube channels, kids songs and stories, built on this exact pipeline. That is the credibility behind the product: it was shaped by actually publishing videos with it, not by guessing what creators need. Decisions like sequential character generation, word-level scene timing and a transparent credit ledger come straight from the friction of producing multilingual content at volume.
You can see the kind of work the pipeline produces in our pages on AI kids song videos and story to video, or browse real published output in the community gallery, where each video shows the recipe that made it. The full map of formats, music, stories, education and more, is on the use cases page.
Which AI models does TubeTube use?
TubeTube doesn't bet on one model. It routes each step to a model suited to that step, and lets you pick the video engine per project. Here's what handles each part of a run:
| Step | Model | What it does |
|---|---|---|
| Music | Suno (V6) | Composes and sings from lyrics you write, or re-records a track you upload as a new sung version (cover). You can also upload a finished mp3 and keep it unchanged. |
| Narration and voices | ElevenLabs | The full ElevenLabs voice library, filterable by language, gender and age, with audio previews. |
| Scene images | GPT Image 2.5, Google Gemini 3 Pro, Flux 2 Pro, Seedream 5 Lite | One image per scene, generated with the previous scenes as context so characters hold. 9 options in the app; the default is GPT Image 2.5 Low, and GPT Image 2.5 comes in four finish levels, from Low to Max. |
| Video (default) | Kling 2.6 Pro | Animates each scene into a silent clip; the song or narration is laid over it in the edit. Kling 2.5 and Kling 3 are also available; Kling 3 can also speak lines on camera in dialogue mode (English, Chinese, Spanish). |
| Video (alternatives) | Google Veo 3.1, Gemini Omni Flash, Seedance 2, Hailuo 2.3 | Other engines you can pick per project, at 720p or 1080p. Veo 3.1 and Gemini Omni Flash can speak lines on camera in dialogue mode; Seedance 2 and Hailuo 2.3 render silent clips. |
Video runs at 720p or 1080p. The point of keeping the same character across scenes is its own topic, see how TubeTube handles AI character consistency with sequential, context-aware generation, reference images, and a transparency report when a scene needs a retry.
How does pricing and transparency work?
TubeTube is credit-based, with a credit worth about $0.011 at the base rate. A 2 to 3 minute video costs a median of 807 credits, about $9; half of them land between 685 and 1,129 credits ($7.50-12.50), measured on 246 videos in August and September 2026. Plans are $22, $49, $129 and $399 per month, 10% cheaper on annual billing. There is no permanent free plan. The exact quote is shown before every launch, and a finished video takes about 15 minutes to generate (median).
For music specifically, TubeTube accepts three input modes: write lyrics and Suno composes and sings them, upload your own mp3 and TubeTube builds visuals around it, or upload a track and request a sung cover. More on that in the AI music video generator, and, for a song you already made with Suno, in Suno song to video.
What TubeTube is not
TubeTube (tubetube.io) is an independent AI video generator founded in 2026. It is not affiliated with YouTube or Google, and is unrelated to the open-source ‘TubeTube’ video downloader on GitHub.
Frequently asked questions
What is TubeTube?
TubeTube is an AI video studio, founded in 2026, that turns written words into finished, multi-scene videos automatically. You give it song lyrics or a narration script, and one run produces the audio, a scene-by-scene storyboard, a consistent image per scene, the animation of each scene, and the final assembled edit. Signup is open. In most countries, new accounts get 1,000 credits at signup, valid for 7 days, with no card required; they aren't granted everywhere (in India, a one-off $4 pack of 1,000 credits replaces them).
How is TubeTube different from a single AI image or video tool?
Most tools do one step. TubeTube runs the whole production in one pass: words to audio, audio to a word-by-word timed storyboard, storyboard to consistent visuals, visuals to animation, then automatic assembly. It also keeps the same character and world across every scene and can dub a narrated video into up to 5 languages at once.
Who is behind TubeTube?
TubeTube was founded by Arsène Huot, a French creator who runs real multilingual faceless YouTube channels (kids songs and stories) built on the exact same pipeline. The product comes from operating those channels day to day, not from a lab, which is the first-party operator experience behind every design choice.
What does TubeTube cost?
Pricing is credit-based, with a credit worth about $0.011 at the base rate. A 2-3 minute video costs a median 807 credits, about $9 at the base rate of $0.011 per credit; half of them cost between 685 and 1,129 credits ($7.50-12.50), measured on 246 videos in August-September 2026. The exact quote is shown before launch and unused credits are refunded. Plans are $22, $49, $129 and $399 per month, 10% cheaper on annual billing, and there is no permanent free plan.
Is TubeTube related to YouTube?
TubeTube (tubetube.io) is an independent AI video generator founded in 2026. It is not affiliated with YouTube or Google, and is unrelated to the open-source 'TubeTube' video downloader on GitHub.
Is TubeTube available to everyone?
Yes, signup is open and creating an account is free. In most countries, new accounts get 1,000 credits at signup, valid for 7 days, with no card required; they aren't granted everywhere (in India, a one-off $4 pack of 1,000 credits replaces them). The credit ledger is transparent and refunds any credits that were reserved but not used.
How long a video can TubeTube make?
Length follows your audio. A narration is read at roughly 14 characters per second, so about 2,500 characters of script make a 3-minute video; a song lasts as long as Suno sings your lyrics, and an uploaded mp3 keeps its own length. Longer audio simply produces more scenes, and the audio is transcribed word by word so scene changes land on the phrasing of the song or narration.
How long does TubeTube take to make a video?
A finished video takes about 15 minutes to generate (median of 984 videos, September 2026); a quarter finish in under 10 minutes and nine in ten in under 32. Every step, from the audio to the final edit, is visible live while it runs, and you can leave the page: the video keeps generating on TubeTube's servers.