If you are asking, "why wont chatgpt create images," the problem is usually not your prompt alone. Image generation can fail because of a temporary service issue, an unstable connection, browser or app interference, an account/workspace setting, or a request that cannot be fulfilled under safety rules.
ChatGPT Images can create new pictures, edit uploaded images, add text, change backgrounds, and work in different aspect ratios. It is available on the web, iOS, and Android, and OpenAI currently states that ChatGPT Images 2.0 is available across tiers. However, access and performance can still vary by account, workspace, device, and current service conditions. OpenAI's Images in ChatGPT guide is the most reliable starting point when the interface or feature availability changes.
Key Takeaways
- Ask directly for an image or use the More > Images option in ChatGPT.
- A long wait, blank result, or generic error can be caused by a temporary service or network problem.
- Disable VPNs, proxies, ad blockers, secure DNS tools, and browser extensions if image requests fail repeatedly.
- Start a fresh chat and use a clear, specific prompt with subject, style, setting, and aspect ratio.
- A refusal is different from a technical error: revise requests that involve unsafe, deceptive, non-consensual, or restricted content.
- If the problem occurs on every device and network, check OpenAI's status page and contact support with useful details.
How Image Creation Works in ChatGPT
You can generate an image in two straightforward ways:
- Type a request in a ChatGPT conversation, such as: "Create an illustration of a rainy evening in Mumbai, with yellow taxis and warm street lights."
- Select More, choose Images, and enter your image prompt.
ChatGPT can also edit an existing image. Upload an image or select one you generated earlier, then explain what should change. For a focused edit, select the area you want to modify and describe the replacement. OpenAI notes that selections are not always perfectly precise, so edits may affect a little more than the highlighted region.
Image generation may take a few minutes for complex requests. Avoid assuming that a short delay means the feature is broken. You can continue using ChatGPT while the image is processing.
First, Identify What "Won't Create Images" Means
The fastest fix depends on the exact symptom. A missing Images option is not the same issue as an endless spinner or a policy refusal.
| What you see | Likely cause | Best first action |
|---|---|---|
| No Images option | Interface, account, workspace, or app issue | Refresh, update the app, start a new chat, and try the web version |
| Endless "Generating" or "Thinking" | Temporary service, connection, or overloaded conversation | Wait briefly, stop generation, then regenerate in a new chat |
| "Something went wrong" | Local browser/network issue or temporary server issue | Check status, use private browsing, disable extensions and VPN |
| "Network error" | Unstable connection, proxy, VPN, or security filtering | Change network and turn off VPN/proxy tools |
| Image request is refused | Safety or policy restriction | Rewrite the request with a safe, consent-based purpose |
| Image is created but looks wrong | Prompt lacks detail or edit selection is too broad | Give more precise visual instructions and regenerate |
This distinction prevents wasted time. For example, repeatedly rewriting a prompt will not solve a blocked network connection, while clearing browser cookies will not override a safety restriction.
Common Technical Problems and Practical Fixes
The image is stuck generating
If ChatGPT stays on "Thinking," "Generating," or an endless spinner, wait around 30 to 60 seconds first. OpenAI recommends stopping generation and choosing Regenerate if the response does not complete.
Then try this sequence:
- Start a new conversation and resend a shorter version of the prompt.
- Refresh the browser or restart the mobile app.
- Sign out of ChatGPT and sign in again.
- Try an incognito/private browser window.
- Switch to another browser, device, or network.
Long conversations with many uploads, instructions, or earlier messages can become less responsive. A fresh chat is often a sensible test because it removes that extra context without requiring you to change your request.
"Something went wrong" appears
This general error can indicate either a temporary OpenAI-side problem or an issue in your local setup. OpenAI's troubleshooting guidance recommends refreshing the page, clearing cache and cookies, using a private window, and disabling extensions—especially privacy and security extensions.
Before making bigger changes, check the OpenAI Status page. A status incident means retrying later may be more useful than repeatedly changing your browser settings. Even when the public status page shows normal operation, individual availability can differ by subscription tier, model, and feature.
Network errors on Indian mobile or office connections
A network error can happen on any connection, but it is particularly worth testing when you switch between mobile data, home broadband, office Wi-Fi, college Wi-Fi, or a public hotspot.
Try these practical checks:
- Turn off a VPN or proxy temporarily.
- Disable secure DNS, web-protection, or network-filtering tools.
- Test the same request on mobile data if Wi-Fi fails, or use Wi-Fi if mobile data is unstable.
- Use a different browser or private browsing mode.
- Avoid frequently switching networks while an image is being generated.
Corporate, educational, and shared networks may use filtering tools that interfere with live connections. If ChatGPT works on mobile data but not on office or campus Wi-Fi, the network configuration is the likely cause.
Browser extensions interfere with ChatGPT
Ad blockers, script blockers, privacy extensions, antivirus web filters, and content-modifying extensions can prevent parts of ChatGPT from loading correctly. This may make the Images option disappear, leave the page blank, or interrupt generation.
The cleanest test is opening ChatGPT in an incognito/private window with extensions disabled. If image generation works there, re-enable extensions one at a time to identify the conflict. Do not disable security software permanently; treat this as a short diagnostic step.
Check Access, App Version, and Workspace Controls
Do not assume an image feature is unavailable simply because you cannot see it in one chat. Try starting a new conversation and look under More for Images. You can also ask ChatGPT directly to create an image.
OpenAI currently says ChatGPT Images is available on web, iOS, and Android. If you are using the mobile app and do not see the expected controls:
- Update ChatGPT through the Apple App Store or Google Play Store.
- Force-close and reopen the app.
- Sign out and sign back in.
- Try ChatGPT in a supported web browser.
For managed workspaces such as Business, Enterprise, Edu, or an organisation-provided account, administrators may control access to some features. If personal ChatGPT works but your work account does not, ask the workspace administrator whether image-related capabilities are enabled for your account.
Avoid relying on old tutorials that tell you to select a specific model name or separate image tool. The ChatGPT interface and feature labels can change over time. Use the current in-product menu and official help documentation rather than screenshots from outdated guides.
Improve Prompts When the Image Is Poor or Incomplete
Sometimes ChatGPT does create an image, but it does not match what you intended. That is a prompt-design issue, not an image-generation failure.
A reliable image prompt includes:
- Subject: what should appear
- Setting: where it is happening
- Style: photo, watercolour, flat vector, editorial illustration, and so on
- Composition: close-up, overhead view, wide shot, centred subject
- Lighting and mood: morning sunlight, dramatic shadows, festive, minimal
- Output format: square, portrait, landscape, or a stated aspect ratio
- Exclusions: what should not appear
Instead of writing:
Make a poster for a café.
Try:
Create a clean vertical promotional poster for an independent Bengaluru coffee shop. Show a ceramic cup of filter coffee on a wooden table, soft morning light, modern Indian editorial illustration style, warm brown and cream palette, space at the top for a headline, no logos, no watermark.
The second prompt gives the system enough direction to make useful choices without being overloaded with contradictory details.
Tips for editing an uploaded image
When editing, state both the area and the required change:
Replace the background behind the product with a light grey studio backdrop. Keep the bottle's shape, label, and lighting unchanged.
If you need to change one small region, use the selection tool and mention the selected area in the instruction. If the edit changes too much, undo it and try a more limited instruction.
When ChatGPT Refuses to Create an Image
A refusal is not necessarily an error. OpenAI applies usage policies designed to prevent harmful, illegal, exploitative, deceptive, or privacy-violating uses.
For example, requests may be restricted if they involve:
- Sexual or exploitative content involving anyone under 18
- Non-consensual intimate imagery
- Harmful threats, harassment, or violent wrongdoing
- Fraud, scams, impersonation, or deceptive identity use
- Using a real person's likeness in a way that could confuse people about authenticity without consent
- Attempts to bypass safety safeguards
These restrictions are explained in OpenAI's Usage Policies. Do not attempt to evade them through indirect wording or repeated retries.
Instead, revise the goal into a safe request. For instance, rather than asking for a deceptive image of a real person, request a clearly fictional character, an original illustration, or a consent-based portrait concept. If you believe a harmless request was declined by mistake, simplify the wording and remove ambiguous details before trying again.
A Step-by-Step Recovery Checklist
Use this order when ChatGPT will not create images:
- Confirm the request is safe, lawful, and does not involve impersonation or non-consensual likeness use.
- Wait briefly if the image is still processing.
- Click Stop generating and then Regenerate.
- Start a new chat and send a shorter, clearer prompt.
- Check OpenAI Status.
- Refresh the page or restart the app.
- Try private/incognito browsing with extensions disabled.
- Turn off VPN, proxy, secure DNS, and web-filtering tools temporarily.
- Try another browser, device, or network.
- Update the ChatGPT mobile app if you are using iOS or Android.
If the issue persists across different browsers, devices, and networks, contact OpenAI Support. Include the approximate time of the failure, the device and browser/app version, the error message, the model used if shown, and whether the same prompt fails in a new chat. Avoid sharing sensitive information in screenshots or support messages.
Frequently Asked Questions
Why won't ChatGPT create images even though I can use ChatGPT normally?
Text chat and image generation can fail differently. The image request may be affected by a temporary service issue, a connection problem, a blocked script, a workspace setting, or a request-specific safety restriction. Test a simple safe prompt in a new chat before changing your account settings.
Can I create images in ChatGPT on a free account?
OpenAI's current help documentation states that ChatGPT Images 2.0 is available on all tiers. Actual availability can still depend on the account, platform, workspace controls, and rollout conditions. If you cannot see it, test the web version, update the app, and start a new chat.
Why does ChatGPT say "Something went wrong" while generating an image?
The error can result from a temporary server issue or a local setup problem. Refresh the page, check OpenAI's status page, clear browser data if needed, use a private window, disable extensions, and turn off VPN or proxy tools.
Why is ChatGPT refusing my image prompt?
The request may conflict with safety rules, especially around minors, non-consensual intimate content, deception, fraud, harmful wrongdoing, or misuse of a real person's likeness. Reframe the request as fictional, consent-based, non-deceptive, and safe.
Why does the edited image change more than the selected area?
OpenAI notes that image-edit selections are not always exact, and edits can extend beyond the highlighted area. Use a smaller selection, explain what must remain unchanged, and retry with a more specific edit instruction.
Conclusion
When ChatGPT will not create images, begin by identifying whether you are facing a technical failure, access issue, weak prompt, or policy refusal. Check the current service status, try a fresh chat, remove browser and network interference, and use a clear prompt with specific visual instructions.
If the problem continues across devices and connections, capture the error details and contact OpenAI Support. A systematic approach is far more effective than repeatedly sending the same prompt—and it helps you get back to creating useful images faster.