No—not in ChatGPT. OpenAI retired GPT-4o from ChatGPT on February 13, 2026. API access is separate: OpenAI’s API model page still lists the gpt-4o alias, but marks GPT-4o as Deprecated. The page lists rates for the model, but the exact shutdown date for a particular API identifier should be checked before you rely on it. (help.openai.com)
ChatGPT’s Free tier is a separate option. OpenAI says the default model and available limits can change over time; free ChatGPT access does not mean GPT-4o is available there. (help.openai.com)
Is GPT-4o free in ChatGPT?
No. GPT-4o was retired from ChatGPT on February 13, 2026. OpenAI’s retirement notice said it would be fully retired across ChatGPT plans after April 3, 2026. ChatGPT still has a Free tier, but the models and limits depend on OpenAI’s current offering. (help.openai.com)
Bottom line: You can use ChatGPT on its Free tier, but GPT-4o is no longer available in ChatGPT. A free account does not mean every model or feature is free or available. (help.openai.com)
Does free ChatGPT access include free GPT-4o API calls?
No. ChatGPT access and API billing are separate. OpenAI’s GPT-4o API documentation lists token rates and says the API Free tier is not supported for the model. It also marks GPT-4o as Deprecated. The model page lists the gpt-4o alias and the gpt-4o-2024-08-06 and gpt-4o-2024-11-20 snapshots as deprecated; it marks gpt-4o-2024-05-13 as shut down. So GPT-4o remains documented as an API option, but it is in the retirement process—not a sound default for a new production integration. (developers.openai.com)
The documented GPT-4o rates are: (developers.openai.com)
| GPT-4o API usage | Price per 1 million tokens |
|---|---|
| Input tokens | $2.50 |
| Cached input tokens | $1.25 |
| Output tokens | $10.00 |
These are API rates in U.S. dollars per million tokens—not ChatGPT subscription fees. Before budgeting or deploying, check the exact identifier’s current status and shutdown date. OpenAI says the model’s shutdown_date can be checked through the API’s List models endpoint. (developers.openai.com)
Example: estimate a text API workload
For illustration, suppose a service makes 10,000 requests in a month, averaging 1,000 input tokens and 500 output tokens per request. If the GPT-4o rates above apply, with no cached input and no additional tool or modality charges:
- Input: 10,000 × 1,000 = 10 million tokens → 10 × $2.50 = $25
- Output: 10,000 × 500 = 5 million tokens → 5 × $10 = $50
- Estimated total: $75
This is a conditional estimate, not a measured bill or confirmation that a particular GPT-4o identifier will remain available for your service. Because GPT-4o is deprecated, treat the calculation as an illustration for evaluating an existing integration—not as a recommendation to build a new one around GPT-4o. Replace the assumed token counts with representative usage and verify the selected model’s rates and shutdown date. Image inputs or other features may affect a real workload’s cost, so this text-only example is not a complete estimate for multimodal traffic. (developers.openai.com)
Should you use GPT-4o for a production system?
OpenAI’s GPT-4o API documentation describes text and image inputs, text output, a 128,000-token context window, and support for function calling and structured outputs. It marks GPT-4o as deprecated, so those capabilities do not by themselves make it a suitable choice for a new production system. (developers.openai.com)
For an existing integration, check the exact model identifier’s status and shutdown date; OpenAI documents a shutdown_date field for that purpose. For a new integration, choose a currently supported model, then test it against your workload and evaluation criteria and estimate cost from expected input and output token volumes. The information cited here does not establish a specific replacement model for your workload. (developers.openai.com)
| Your need | Practical choice |
|---|---|
| Free general-purpose chatting | Use ChatGPT’s Free tier and the models currently offered there; GPT-4o is retired from ChatGPT. |
| GPT-4o specifically in ChatGPT | Not available: OpenAI retired it from ChatGPT in February 2026. |
| GPT-4o API for an existing integration | The API documentation lists GPT-4o as deprecated. Check the exact identifier’s current status and shutdown date before relying on it. |
| A new production integration | Evaluate a currently supported API model against your capability, reliability, and budget requirements rather than choosing GPT-4o solely because you recognize its name. |
FAQ
Is GPT-4o free for ChatGPT Free users?
No. OpenAI retired GPT-4o from ChatGPT on February 13, 2026. The ChatGPT Free tier remains available, but its models and limits depend on OpenAI’s current offering. (help.openai.com)
Is GPT-4o API usage free?
The GPT-4o API documentation lists paid token rates and says the API Free tier is not supported for the model. GPT-4o is marked deprecated, so check the exact identifier’s current status before relying on it. (developers.openai.com)
Is GPT-4o still available through the API?
OpenAI’s model page still lists the gpt-4o alias, while marking GPT-4o as deprecated. The exact shutdown date for a specific identifier should be checked through the model’s shutdown_date field in the API. Don’t assume a deprecated model will remain available for a new or continuing production service. (developers.openai.com)
Does ChatGPT Plus include API credits?
Do not assume so. ChatGPT and API billing are separate. Check the terms of any applicable API-credit benefit and the credit balance in the linked API organization before including credits in a budget.
Sources
Retiring GPT-4o and other ChatGPT models — OpenAI Help Center