Paid · account required to generate
Cutshed Studio
Generate a video, an image, a face swap, an emoji, or audio, and pay for that job in credits. References are sent only when the model accepts them. This is the only paid part of Cutshed. The editors on the rest of the site stay free and never see this account.
Packs are $5 for 400, $10 for 850, and $25 for 2,250. A 5-second Seedance 2.0 Mini clip at 480p costs 30 credits. The rest of the rate card is on the pricing page.
Checking Studio…
Models
These are the models in the catalog. The picker shows the ones the current account can run. A price is never typed into the page: it is quoted from the catalog when you choose a length and a resolution.
Video
- Seedance 2.0 Mini. The default. Text or a start frame, with optional synced audio.
- Seedance 2.0 Fast. The faster Seedance 2.0 pass, 480p or 720p.
- Seedance 2.0. Full Seedance 2.0, from 480p through 4K.
- Seedance 2.5. Newest Seedance. Clips up to 30 seconds at 480p or 720p.
- Wan 3.0. Alibaba Wan 3.0. Text or a start frame, 2 to 30 seconds.
- Wan 3.0 Prime. Higher Wan 3.0 tier. Text or a start frame, 2 to 30 seconds.
- Grok Imagine Video. Text or a start frame. Audio is always on. One rate for 480p and 720p.
- Grok Imagine Video 1.5. Newest Grok preview. A start image is required. Audio is always on.
- Gemini Omni 1.1 Flash. Google’s Omni video model. It picks the length. Priced here as 10 seconds, which is what a typical clip runs.
- Veo 3.1. Google Veo 3.1. 4, 6, or 8 seconds, with audio on or off.
- Veo 3.1 Fast. Faster Veo 3.1. Same lengths, lower price, audio on or off.
- Veo 3.1 Lite. Lowest-cost Veo 3.1. Audio is always on. 1080p is 8 seconds only.
- Kling 2.5 Turbo Pro. Fast Kling turbo. 5 or 10 seconds, optional start frame.
- Kling 2.6. Kling 2.6 with optional native audio. 5 or 10 seconds.
- Kling 3.0. Newest Kling. Standard is 720p, Pro is 1080p, plus 4K. Audio is optional.
- Kling 3.0 Omni. Kling 3.0 Omni. Same lengths and resolutions, with a different audio price.
- MiniMax H3. Text, a first and last frame, or up to 9 images, 3 videos, and 3 audio clips. 768P or 2K, from 4 to 15 seconds.
- Hailuo 2.3. Text or an optional start frame. 6 or 10 seconds at 768p. 1080p is 6 seconds only.
- Hailuo 2.3 Fast. Needs a start image. Same lengths as Hailuo 2.3, at a lower price per clip.
Image
- Nano Banana 2.1. Latest Nano Banana. Text or up to 14 reference images, at 1K, 2K, or 4K.
- Nano Banana Pro. Nano Banana Pro. The same kind of reference edit, at a higher 1K, 2K, and 4K price.
- Nano Banana. The original Gemini 2.5 image model. One price, with reference edits.
- Seedream 5.0 Pro. ByteDance Seedream 5.0 Pro. Text or up to 10 reference images, at 1K or 2K.
- Seedream 5.0 Lite. Seedream 5.0 Lite. Up to 14 references, and up to 4 related images when you ask for more than one.
- FLUX.2 Pro. FLUX.2 Pro. Up to 8 reference images. The price is the run plus the output megapixels.
- FLUX.2 Dev. FLUX.2 Dev with the fast setting left on. Quoted as about 1 megapixel, with up to 5 reference images.
- FLUX.1 Kontext Pro. FLUX.1 Kontext Pro edits one image from a text instruction.
- FLUX.1 Schnell. FLUX.1 Schnell. Up to 4 images a run. It does not take reference images.
- Imagen 4 Ultra. Imagen 4 Ultra. Text to image at 1K or 2K, at the same price.
- Imagen 4. Google Imagen 4. Text to image at 1K or 2K, at the same price.
- Imagen 4 Fast. The faster Imagen 4 pass. Text only.
- Grok Imagine Image Quality. Higher-detail Grok image at 1k or 2k. One image can be edited.
- Grok Imagine Image. xAI Grok image. Text, or an edit of one image.
- Ideogram v4 Quality. Ideogram 4.0 at the quality setting. Text to image. Pick a pixel size, or let Ideogram choose.
- Ideogram v4 Balanced. Ideogram 4.0 balanced. Text to image. Pick a pixel size, or let Ideogram choose.
- Ideogram v4 Turbo. Ideogram 4.0 Turbo. Text to image. Pick a pixel size, or let Ideogram choose.
- Recraft V4 Pro. Recraft V4 Pro. Higher resolution, same design-focused text-to-image model.
- Recraft V4. Recraft V4. Text to image at standard resolution, with an aspect ratio.
Face swap
- Face swap. Puts a source face onto a target image. Consent is required.
- Face swap video. Replaces the person in a target video with one source face. Priced per second of that video.
- Recast (MiniMax H3 Max). Replaces the people in a source video with up to four reference photos, and keeps the original motion, cuts, and audio.
Emoji
- Emoji maker. A square emoji with a transparent background. One to four variations. A face photo uses FLUX.2 Dev so the emoji can follow that face; otherwise FLUX.1 Schnell draws it. Background removal is 851-labs/background-remover.
Music
- MiniMax Music 2.0. A full song with vocals for $0.03. Lyrics are required. Length is chosen by the model.
- Google Lyria 2. Google's music model. One 30-second clip for $0.10. Best as an instrumental bed.
- MiniMax Music 2.5. Higher-fidelity song with singing, or an instrumental if you leave the lyrics blank. $0.15 a track.
Speech
- ElevenLabs Turbo. Fast ElevenLabs speech at $0.05 per 1,000 characters. Speed and stability are both available.
- MiniMax Speech HD. MiniMax Speech-02 HD at $0.10 per 1,000 characters, with speed control and a system-voice list.
- Kokoro. The cheaper voice. Apache-2.0 Kokoro at $0.02 per 1,000 characters. American English, with speed control.
Sound effects
- ElevenLabs Sound Effects. Text to a sound effect at $0.002 per second, from 2 to 22 seconds.
FAQ
Do I need an account to generate?
Yes. Studio asks you to sign in before a job starts. A new account starts at zero credits. The free editors and tools on the rest of the site do not use this account, and they do not upload the file you open.
How much does a clip cost?
Packs are $5 for 400, $10 for 850, and $25 for 2,250. A 5-second Seedance 2.0 Mini clip at 480p is 30 credits. Every other rate is on the pricing page, taken from the same catalog. Credits stay on the account until you spend them.
What if the model fails or blocks the prompt?
The credits go back. A pack you do not use is not refunded for cash, except where the law requires it. Studio is not a subscription.
Does Studio upload my file?
A generation sends the prompt, and a reference only when that model accepts one, to the provider that runs the model. The free tools never do that. If a finished file is not stored on Cutshed, download it within an hour.
Steps
- Sign in. New accounts start at zero credits.
- Buy a pack on the pricing page if the balance is short.
- Pick Video, Image, Face swap, Emoji, or Audio, then a model.
- Write the prompt, add only the references that model accepts, and generate. A failed or blocked job returns the credits.