Start creating

Models for video, images, voice and music

The world's best models. One studio.

Make films, photos, voices and music with Seedance 2.5, Veo, Kling, Nano Banana, ElevenLabs and every other top model, side by side in one studio. Find the one that fits your idea, then make it yours.

  • No subscription
  • Every top model, one account
  • Pay per run, priced before it starts
What you want to make
What it does

Video

Make it move

Direct a scene, animate a still, film a product. The models behind the clips people share.

  • Nine years, made on Leap

    Veo 3.1

    Google

    Google's highest-quality video model, with sound.

    • Text to video
    • Image to video
    Try it
  • Soba under the tracks, made on Leap

    Sharp, steady video at a very low price, with optional sound.

    • Text to video
    • Image to video
    Try it
  • Rooftop leap, made on Leap with Seedance 2.5

    Seedance 2.5

    ByteDance

    Lifelike motion with sound made in the same take, from text, a first frame with no real person in it, or up to 30 images, 10 videos and 10 sounds it follows.

    • Text to video
    • Image to video
    Try it
  • Northbound, shot 3, made on Leap

    MiniMax H3

    MiniMax

    MiniMax's 2K video model, with sound.

    • Text to video
    • Image to video
    Try it
  • Chobe crossing, made on Leap

    Wan 3.0

    Alibaba

    Alibaba's newest video model, with sound, from text or a first frame.

    • Text to video
    • Image to video
    Try it
  • Riad, made on Leap

    FLUX.3 Video

    Black Forest Labs

    Black Forest Labs' first video model, from text or a first frame.

    • Text to video
    • Image to video
    Try it

More video models

  • Wan 2.6Alibaba

    Open-weight video with a soundtrack, at 720p or 1080p.

    Try
  • Seedance 2.0ByteDance

    ByteDance's reliable all-rounder: lifelike motion with sound, from text or a first frame with no real person in it.

    Try
  • Seedance 2.0, faster and for less, up to 720p: quick takes with sound, from text or a first frame with no real person in it.

    Try
  • The lowest-priced Seedance 2.0: clips with sound up to 720p, for drafts and volume, from text or a first frame with no real person in it.

    Try
  • Moves the character in your image the way the person in a reference video moves, and keeps the face through turns, hands and hats. Billed by the length of the reference video.

    Try
  • Fast video with sound, from text or a first frame.

    Try

Image

Make it real

Posters, products, portraits and whole worlds, from one sentence or your own photo.

  • Pav bhaji at midnight, the generated first frame

    FLUX.2 [max]

    Black Forest Labs

    The highest-quality FLUX.2, for hero images.

    • Text to image
    Try it
  • The first frame of 5 da manhã, shot 1

    Seedream 4.5

    ByteDance

    2K or 4K images with crisp typography, and edits of up to 14 reference images.

    • Text to image
    • Image edit
    Try it
  • The first frame of The tailor

    ByteDance's newest: detailed images up to 2K and edits of up to 10 reference images.

    • Text to image
    • Image edit
    Try it
  • The first frame of Dawn run, Iten

    xAI's second Grok Imagine, with a 2K option and a cheaper draft quality.

    • Text to image
    Try it
  • The first frame of Burner

    Gemini 3 Pro Image: Google's best for legible text, infographics and edits of up to 6 references.

    • Text to image
    • Image edit
    Try it

One prompt, 6 models

“A hand-painted shop sign that reads OPEN LATE above a tiny ramen bar on a rainy night, warm light spilling onto wet cobblestones, cinematic 35mm photo”

More image models

  • FLUX.1 [schnell]Black Forest Labs

    Fast, low-cost text to image. Good for drafts and volume.

    Try
  • FLUX.2 [pro]Black Forest Labs

    Black Forest Labs' second generation: sharper detail and text than FLUX1.1, at a lower price.

    Try
  • FLUX.2 [flex]Black Forest Labs

    FLUX.2 tuned for typography and fine detail, such as posters and layouts.

    Try
  • FLUX.2 [klein] 9BBlack Forest Labs

    A small, fast FLUX.2, a step up in quality from klein 4B.

    Try
  • FLUX.2 [klein] 4BBlack Forest Labs

    The smallest, fastest FLUX.2, for volume.

    Try
  • Seedream 4.0ByteDance

    The lowest-priced Seedream: 1K to 4K images and edits of up to 14 reference images.

    Try

Voice and sound

Give it a voice

Narration, characters, songs and sound effects for everything you make. Press play.

  • TTS-1

    OpenAI

    0 / 7 s
    Read the transcript

    Hi, I'm testing Leap's voices. Can you hear the difference between each one? Let's find out together.

    OpenAI's fast text to speech, in nine voices.

    • Text to speech
    Try it
  • TTS-1 HD

    OpenAI

    0 / 7 s
    Read the transcript

    Hi, I'm testing Leap's voices. Can you hear the difference between each one? Let's find out together.

    OpenAI's higher-quality text to speech, in nine voices.

    • Text to speech
    Try it
  • 0 / 7 s
    Read the transcript

    Hi, I'm testing Leap's voices. Can you hear the difference between each one? Let's find out together.

    Expressive speech in five voices.

    • Text to speech
    Try it
  • MAI-Voice-2.1

    Microsoft

    0 / 8 s
    Read the transcript

    Hi, I'm testing Leap's voices. Can you hear the difference between each one? Let's find out together.

    Microsoft's most natural voices, for narration, across 23 languages.

    • Text to speech
    Try it
  • 0 / 8 s
    Read the transcript

    Hi, I'm testing Leap's voices. Can you hear the difference between each one? Let's find out together.

    The fast MAI voice, at a lower price.

    • Text to speech
    Try it
  • ACE-Step

    ACE Studio

    0 / 60 s

    Open-weight songs from genre tags and lyrics, priced by the second.

    • Music
    Try it

More voice and sound models

  • Gemini's lighter speech model: 30 voices, directed in plain words.

    Try
  • S2.1 ProFish Audio

    Fish Audio's most natural speech. Direct a line with tags in the text, such as [whispering].

    Try
  • S2 ProFish Audio

    Fish Audio's model before S2.1, at the same price and in the same voices.

    Try
  • Songs up to five minutes long from a style prompt and your lyrics. Without lyrics it makes an instrumental. A song can end before the length you ask for, which is what the run is billed.

    Try

3D

Make it solid

Turn a picture or a sentence into a 3D model you can spin, print or drop into a game.

  • RodinDeemos

    Production-ready 3D assets from an image or a prompt.

    Try
  • Rodin 2.5 in seconds, at ten cents a model.

    Try
  • Meshy 6Meshy

    Detailed, textured 3D models from an image.

    Try
  • Detailed, textured 3D models from a prompt.

    Try
  • TRELLISMicrosoft

    Open-weight image to 3D, at two cents a model.

    Try
  • A textured 3D model from one image, as a GLB. Without texture, a plain white mesh costs a third as much.

    Try

Transcripts

Read every word

Transcripts, captions and subtitles from any recording, timed to the word and labelled by speaker.

  • Scribe v2ElevenLabs

    ElevenLabs' most accurate transcription: word timings, speaker labels and 90+ languages, from a sound or a video. Billed per started minute of the recording.

    Try
  • OpenAI's Whisper on fal's fast runtime: the cheapest way to transcribe hours of audio, in 99 languages, and it translates any of them into English. Segment timings, no speaker labels.

    Try

Use them together

One model makes the still. Another makes it move.

"5 da manhã, shot 4", replayed: Nano Banana Pro makes the still, then FLUX.3 Video brings it to life, in the same studio and the same library.

The first frame of 5 da manhã, shot 4
5 da manhã, shot 4, made on Leap
  1. 01Nano Banana ProMakes the stillDone

    Low-angle wide shot, 24mm, dawn: the dancer from the reference image, same clothes, on the terrace of a Lisbon miradouro, terracotta rooftops and the...

  2. 02FLUX.3 VideoBrings it to lifeDone

    He breaks the freeze into a fluid wave through his arms and body as the sun touches the river behind him; pigeons lift off the balustrade. Slow dolly...

  3. "5 da manhã, shot 4", made on Leap: two models, one studio, one library.

Why Leap

Why creators make it on Leap

  1. 01

    Every top model, side by side

    173 models from Black Forest Labs, Google, ByteDance, Kuaishou, Alibaba and more, side by side in one studio.

    See the models
  2. 02

    One click, or full control

    Character Studio, effects and recipes write the prompt for you. Or write your own for any model.

    See the tools
  3. 03

    Your work stays yours

    Everything you make stays in your library while your account is active, and you own what you make here.

    See what people make
  4. 04

    One balance for everything

    The studio, the API and your agents spend the same balance, and every run shows its price before it starts.

    How pricing works