o4-mini Models by OpenAI
Last refreshed 2026-10-17. Next refresh: weekly.
Capabilities
About
o4-mini is OpenAI’s small o-series reasoning model in the API, aimed at fast, efficient reasoning with strong coding and visual-task performance. It takes text and image input and returns text, supports reasoning tokens, function calling, structured outputs, streaming, and fine-tuning, and offers a large context window with high max output length. OpenAI documents it as succeeded by GPT-5 Mini.
Decision facts
Archived Variants
Use-when guidance is based on each model's tracked capabilities, context window, release date, and replacement status.
Keep for legacy integrations; evaluate GPT-5.6 Terra before new work.
Keep only for existing workloads; choose a current variant for new builds.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| o4-mini | Keep for legacy integrations; evaluate GPT-5.6 Terra before new work. | 2025-04 | 200K contextreasoningJSON / Tool use | Replaced |
| o4 Mini Deep Research | Keep only for existing workloads; choose a current variant for new builds. | 2025-04 | 200K contextstructured outputsmultimodal inputs | Archived |
Release Timeline
1 release groupReplaced By
Keep for legacy integrations; evaluate GPT-5.6 Terra before new work.
Specifications(2 models)
| Model | Released | Context | Vision | Multimodal | Reasoning | JSON / Tool use | Structured Outputs | Code Exec |
|---|---|---|---|---|---|---|---|---|
| o4-mini | 2025-04 | 200K | Yes | Yes | Yes | Yes | Yes | Yes |
| o4 Mini Deep Research | 2025-04 | 200K | Yes | Yes | No | No | Yes | No |
Available From(4 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| o4-mini | Replicate API | $1 | $4 | Serverless |
| o4-mini | OpenAI API | $1.1 | $4.4 | Serverless |
| o4-mini | OpenRouter | $1.1 | $4.4 | Serverless |
| o4-mini | Vercel AI Gateway | $1.1 | $4.4 | Serverless |
| o4 Mini Deep Research | OpenRouter | $2 | $8 | Serverless |
| o4 Mini Deep Research | OpenAI API | $2 | $8 | Serverless |
