Integrations
Uses the OpenAI API for DALL-E 3 image generation and GPT-4 Vision analysis, with the user's own API key. No other integrations are described on the homepage.
Image Recursor
Turn one image prompt into an evolving visual chain.
Beginner API
Own Image Recursor? Verify ownership to get a dofollow link
What is Image Recursor?
Image Recursor is a generative image tool that turns one image prompt into an evolving visual chain. Image Recursor is a web tool that chains DALL-E 3 and GPT-4 Vision to evolve an image across repeated loops, using preset or custom transformation prompts.
Image Recursor is a browser-based tool that generates a chain of images by combining DALL-E 3 and GPT-4 Vision. You begin with a starting image prompt and a number of loops, and the tool repeatedly passes the latest image through GPT-4 Vision with a chosen transformation prompt before sending the result back to DALL-E 3. The effect is a visual sequence that drifts from the original idea, similar to a game of telephone played with pictures. The interface is deliberately minimal. You enter an OpenAI API key, which the site says stays on-device, then choose a starting prompt, a loop count and a transformation style. Preset options include making the image scarier, more random, more mystical, more intense or funnier, and a custom prompt field is available for other directions. Each generate action produces the chain in the browser, so the current state of an image becomes the input for the next round. The tool suits people who want to explore how a visual concept changes under repeated reinterpretation: prompt experimenters, artists looking for unexpected variations, and anyone curious about the behaviour of vision-language models. It is not a production image editor or a batch generator for finished assets. Its value is in the iterative loop itself, and it sits closer to a creative toy or a research sketchpad than to a full design suite. Because it relies on your own OpenAI key, usage costs and rate limits are handled through your OpenAI account rather than a subscription on the site. The homepage does not describe pricing tiers, a support channel or third-party integrations beyond OpenAI. Image Recursor works as a single-page web app with no sign-up flow shown, which keeps setup short but also means the experience is tied to the capabilities and limits of the underlying OpenAI models.
How do you use Image Recursor?
1Open the web app
Load the Image Recursor page in a browser. The whole tool runs on one screen with a starting prompt field, loop count, API key input and transformation selector.
2Enter your OpenAI API key
Paste an OpenAI API key into the key field. The homepage states the key stays on-device, so it is used locally by the page rather than stored on a server.3Write the starting image prompt
Describe the first image you want DALL-E 3 to create. This prompt anchors the chain, and every later image is derived from it through the loop.4Set the loop count
Choose how many iterations the chain should run. Each loop applies the transformation prompt to the current image and feeds the result into the next generation.5Pick or write a transformation prompt
Select a preset such as Make it scarier or Make it funnier, or enter a custom GPT-4 Vision prompt. This prompt decides how the image shifts at each step.6Generate and review the chain
Press Generate and watch the sequence build. Review how the image changes across loops and adjust the prompt or loop count to steer the next run.
Pros and cons
Pros
- Simple loop concept that produces surprising visual drift from a single starting promptAI
- Preset transformation prompts (scarier, funnier, mystical and others) lower the setup effortAI
- API key is described as staying on-device, which reduces exposure concernsAI
- No account creation is shown, so a session can start as soon as a key is enteredAI
- Works entirely in the browser on a single pageAI
Cons
- Requires an OpenAI API key, so there is no way to try it without an OpenAI account and billing setupAI
- Costs are tied to OpenAI usage and the site does not explain or cap themAI
- No pricing, support or documentation pages are provided on the homepageAI
- Output quality depends entirely on GPT-4 Vision and DALL-E 3 behaviour and limitsAI
- No export, history or project management features are describedAI
How much does Image Recursor cost?
Features
Image Recursor chains DALL-E 3 image generation with GPT-4 Vision analysis across a user-defined number of loops. Users supply a starting prompt, a loop count and either a preset transformation prompt (scarier, random, mystical, intense, funnier) or a custom one. The OpenAI API key is entered in the browser and described as staying on-device.
Frequently asked questions about Image Recursor
Does Image Recursor have an API?
Yes, Image Recursor offers an API.How do you use Image Recursor?
The walkthrough on this page covers 6 steps: 1. Open the web app 2. Enter your OpenAI API key 3. Write the starting image prompt 4. Set the loop count 5. Pick or write a transformation prompt 6. Generate and review the chain.What platforms does Image Recursor support?
Image Recursor is available on Web App.What does Image Recursor integrate with?
Uses the OpenAI API for DALL-E 3 image generation and GPT-4 Vision analysis, with the user's own API key. No other integrations are described on the homepage.What are the limitations of Image Recursor?
Requires an OpenAI API key, so there is no way to try it without an OpenAI account and billing setup. Costs are tied to OpenAI usage and the site does not explain or cap them. No pricing, support or documentation pages are provided on the homepage.
Status
StatusActive
Views0
Outbound clicks0
Added10/21/2026
Platforms
Web App
Pricing
Free
