OpenAI’s public Free Tier FAQ does not publish a fixed number of image generations for ChatGPT Free. It says image generation has a separate usage limit and that ChatGPT notifies you when you reach it; the product shows the reset timing. So there is no verified public “X images per day” figure to rely on. (help.openai.com)
To find the limit that applies to your account, use ChatGPT’s image-generation feature and follow the in-product notice if you reach the cap. OpenAI says the reset timing appears in ChatGPT. Its FAQ does not specify a universal daily, weekly, or monthly image allowance. (help.openai.com)
What counts as a generation, and when does it reset?
OpenAI confirms that image generation has its own limit, separate from the main Free-tier chat limit. But its public FAQ does not say whether a rejected result, an edit, or each image in a request for several variations counts as one generation. It also does not publish a general reset interval. The dependable guidance is the limit notice and reset information shown in your ChatGPT account—not a count inferred from older articles or another user’s experience. (help.openai.com)
ChatGPT Images supports both creating and editing images, but that capability description does not specify how those actions are charged against a Free allowance. (help.openai.com)
What does the production image API cost?
The OpenAI API is billed separately from ChatGPT. Using ChatGPT Free does not include API image-generation usage. OpenAI’s API documentation lists GPT Image 2.5 models for generation and editing; it describes Flare as the fast option for everyday work and Sunburst for workflows where editing precision matters. (help.openai.com)
For both models, the published output rate is $30 per 1 million image-output tokens with Standard processing, or $15 per 1 million with Batch processing. Image input tokens are listed at $8 per 1 million; text input is $5 per 1 million. These are token rates, not a guaranteed flat price per generated image: the cost depends on token use, including the image output. (developers.openai.com)
Illustrative budget: if a workload used 100,000 image-output tokens, the output-token charge would be $3 at the Standard rate or $1.50 at the Batch rate, before any input-token charges. That arithmetic is only an example; it does not assume that every image uses the same number of tokens. Check your actual usage and current API pricing before forecasting a production budget. (developers.openai.com)
The API model pages also show rate limits by API usage tier, with access and limits varying by tier. API access is therefore not an unlimited-throughput substitute for a ChatGPT plan allowance. (developers.openai.com)
Which option makes sense?
- Occasional personal images: Start with ChatGPT Free. Its exact allowance is not stated publicly, so use the in-product notice to learn your account’s limit and reset time.
- A predictable production workload: Compare API costs using your expected token usage, model, and processing option. Don’t estimate a flat cost per image unless you have measured or otherwise established the token use for your own outputs.
- Unsure whether the API is included: It is billed separately from ChatGPT. Review API billing and usage independently from your ChatGPT account. (
help.openai.com)
Bottom line: OpenAI does not state a single Free-plan image-generation count in its public FAQ. ChatGPT shows the applicable limit and reset timing in-product; API image generation is a separate, token-priced production option. (help.openai.com)