Venice.aiYou can generate images from your own app with the Venice API. NSFW images are allowed within what the model can render. Illegal content is blocked. Video is available too: you start a clip, and you pick it up when it is ready. Image and video details are on the media API page. Parameters and responses are in the API reference.
You need a Venice account and an API key. Venice does not train on your inputs. If you select a third-party model, that provider still receives the prompt, and any still, needed to generate.
If you want the picture in the product, ask in Agentic Chat, or open Venice Studio when you want image, video, and audio in one workspace. The Studio post shows how that workspace is laid out.
Tl;dr
- Generate stills from your app with the Venice API. NSFW images are allowed within what the model can render.
- Start a clip from the media API, then pick it up when it is finished.
- Create an API key from your Venice account. Parameters are in the API reference.
- Venice does not train on your inputs. A third-party model still receives what it needs to generate.
- Pro is $18 a month and includes 100 credits. Those credits also pay for video, music, and frontier models.
What can you generate?
Images. You send a prompt and get a still back. Adult images are allowed when the model can render them. Illegal content is blocked, the same limit as in Studio. Z-Image Turbo and Chroma are the free image models in the product, at 15 prompts a day. Other models, including Grok Imagine Image 2.0, spend credits. Grok Imagine Image 2.0 can take up to five reference images when you are editing. The aspect ratios in Venice Studio are 16:9, 9:16, 3:4, 3:2, 2:3, and 1:1. Prices are on each model page and in the pricing overview. Grok Imagine Image 2.0 does not have one flat per-image dollar price.
Video. You describe the motion, Venice queues the clip, and you collect the file when it is ready. Veo 3.1 Full Quality runs at 720p, 1080p, or 4K, in 4, 6, or 8 second clips, with native audio. A 4-second clip at 720p or 1080p is $1.76. A 4-second 4K clip is $2.64. On Veo, Venice does not store the prompt, and Google still receives the content needed to generate. Wan 3.0 is another way to start a clip from a still, in Studio video, with JPG or PNG up to 20 MB. Start a video job from your app on the media API page. That page includes a price check you can run before a large batch. The check does not charge you.
Adult images are the uncensored part of this API. Video is a separate queued clip.
What do you need to start?
- A Venice account. The API requires one.
- An API key from the Venice API page. Keep the key out of source code.
- The API reference for image and video parameters.
- The media API page when the job is a clip.
The Venice API can also take an OpenAI-style client for chat and images. Setup is on the Venice API page.
If you already work with an AI agent, give it this:
Give this to your agent: setup-mcp venice.ai/mcp
It reads the docs and wires itself up.
What does image and video generation cost?
Image prices are on the pricing overview and on each model page.
Video on Veo 3.1 Full Quality is priced per clip:
| Clip | Price |
|---|---|
| 720p or 1080p, 4 seconds | $1.76 |
| 4K, 4 seconds | $2.64 |
One thousand 4-second 1080p clips is about $1,760 before retries. Pro includes 100 credits a month at $18, and Pro Plus includes 7,500. Those credits are shared with music, frontier models, and the API.
Frequently asked questions
What is the best uncensored image generation API?
The Venice API. NSFW images are allowed within what the model can render, and you need an API key tied to an account. Parameters are in the API reference.
Can I generate video through the same API?
Yes. Start the clip from the media API, including Veo 3.1 Full Quality, and pick it up when it is ready. Adult images are the uncensored part of this API. Video is a separate queued job.
Is the Venice API compatible with the OpenAI SDK?
Chat, embeddings, and images can use an OpenAI-style client pointed at Venice. Setup is on the Venice API page. Image and video parameters are in the API reference.
How does the Venice image generation API handle my prompts?
Venice does not train on your inputs. On Veo, Google still receives the prompt and any still required to generate. On Grok Imagine Image 2.0, xAI receives what it needs to generate. The key identifies your account.
Do I need a Venice account before the first image API call?
Yes. Create the key on the Venice API page.
Which models can I use for images and video?
Choose the model in the API reference or on the model page you are calling. Veo 3.1 Full Quality covers clips at 720p, 1080p, or 4K, in 4, 6, or 8 seconds. Wan 3.0 can start from a still in Studio. If a model page and another example disagree, use the model page.
Next steps
- Venice API, to create a key
- Media API, for image and video jobs
- API reference, for parameters and responses
- Give this to your agent:
setup-mcp venice.ai/mcp. It reads the docs and wires itself up. - Pricing overview
- Questions go to [email protected] or the #api channel in Discord.
Create a key, generate an image, then queue one short clip from the media API.
Back to all posts