Join the waitlistWaitlist Menu

Models Coming

The models behind your ads. Right inside Claude.

Video, stills, voice, lip-sync and music, through one Turntake connection. Pay with credits as you go. No API keys, and no account at every model maker. Not open yet, so there’s nothing to try today.

Join the waitlist See the lineup

Early access we set it up with you. Coming we’re building it.

How it will work Coming

Connect once. Ask for the shot.

  1. 01

    Connect once.

    Add Turntake to Claude and sign in. No keys to copy.

  2. 02

    Ask for what you need.

    Animate this still. Voice this script. Make a talking head from this portrait. Pick the model yourself, or let Claude suggest one.

  3. 03

    Pay as you go.

    Credits cover each generation, from one balance. You see the cost before you run it.

Under the hood it’s MCP, the standard way Claude plugs into outside tools.

The lineup Coming

Six jobs. The models that do them.

The lineup is a plan and will change. Model names belong to their makers.

Video

Animate a still

B-roll, product moments and motion between scenes, from a still or a prompt.

  • Seedancebytedance/seedanceTurns a still into a short shot. Also works from text, or from a first and last frame.
  • Klingkuaishou/klingShort cinematic clips from text or a still.
  • Veogoogle/veoVideo from text or an image, with generated sound.
  • Hailuominimax/hailuoVideo from text or an image.
Images

Make the still

Start frames, product stills and the same character across shots.

  • Nano Bananagoogle/nano-bananaMakes and edits images by conversation, and keeps a character consistent across shots.
  • Fluxblack-forest-labs/fluxText-to-image, with variants for product shots and clean type.
  • Seedreambytedance/seedreamText-to-image with a cinematic look.
  • GPT Imageopenai/gpt-imageText-to-image that can put readable text in the picture.
Voice

Voice the script

A voiceover for each market, from a script.

  • Eleven v4elevenlabs/eleven-v4Expressive speech from a script, with tags for tone, in many languages.
  • Speech-to-speechelevenlabs/speech-to-speechTurns a recorded read into another voice and keeps the delivery.
Talking heads

Make the face talk

A front-facing portrait and a voice track become a talking video.

  • p-video-avatarprunaai/p-video-avatarTurns a front-facing portrait and a voice track into a talking video.
  • OmniHumanbytedance/omnihumanAudio-driven video of a person from a single image.
  • Lipsyncsync/lipsyncLip-sync for existing video, for dubbing into another language.
Music and sound

Score the ad

A music bed, timed to the length of your ad.

  • Eleven Musicelevenlabs/eleven-musicMusic from a text prompt, timed to the length of your ad.
Speech-to-text

Time the captions

A time for every word, so captions land on the spoken word.

  • Scribeelevenlabs/scribeTranscribes speech with a time for every word.
Join the waitlist

Why through Turntake Coming

Models make shots. Skills make ads.

A model gives you a clip. The factory around it decides which clip, how it’s briefed and whether it’s good enough to ship.

  • No keys. One balance.

    One sign-in and one credit balance, instead of an account and a key at every model maker.

  • Briefed the way we brief them.

    Our skills already write the prompts we use with these models for our own ads. Through Turntake, your Claude will get the same playbook.

  • Library first.

    New shots only for the beats you’re missing. Everything you make will land in your library, so you can use it again.

  • Checked before you see it.

    When the shots become an ad, loudness, safe zones and caption lines will be checked before you see the take.

See the factory Bring your own content

FAQ

Is this available now?

No. It’s coming. Join the waitlist, and we’ll write when there’s something to try.

Which models will be included?

The lineup above is the plan. It will change, and we’ll say so when it does.

What will it cost?

We haven’t set prices yet. We’ll share pricing before launch.

Do I need my own API keys?

You won’t. That’s the point: you’ll sign in to Turntake once, and credits will cover the rest.

Does it work outside Claude?

We’re starting with Claude. Other AI apps aren’t decided yet.

Who owns what I make?

It depends on each model’s terms. We’ll spell that out before anything opens.

Learn more

All guides from our own production

Waitlist

The models, in Claude. Coming next.

Models open in early access. Leave your email, and we’ll write when there’s something to try.

Early access we set it up with you. Coming we’re building it.