9
Models listed
OpenAI Coding Models
Price & capability guide for coding workflows.
A clean single-page reference with model comparisons, hover states, and a mobile-friendly layout built for fast scanning.
Quick take
Use mini models for speed, reasoning models for the hard bugs.
Compare input and output pricing, then match the model to your task.
2
Model groups
1
Fast reference page
Compare
Model Table
Hover rows for emphasis. On mobile, the table scrolls horizontally for easy reading.
| Group (Modality) | Model Name | Input Price (per 1M tokens) |
Output Price (per 1M tokens) |
Coding Tier | Best Use Case |
|---|---|---|---|---|---|
| Multimodal (Accepts Image + Text) | gpt-4o-mini | $0.15 | $0.60 | Light / Scripting | Everyday code modifications, quick scripts, UI design translation from screenshots. |
| Multimodal (Accepts Image + Text) | gpt-5.4-mini | $0.375 | $2.25 | Standard / Production | Rapid execution, multi-file software scaffolding, well-rounded development. |
| Multimodal (Accepts Image + Text) | gpt-4.1 | $2.00 | $8.00 | Advanced / Heavy | High-complexity algorithms, deep codebase analysis, complex system debugging. |
| Multimodal (Accepts Image + Text) | gpt-5.4 | $2.50 | $15.00 | Frontier | State-of-the-art software engineering, database design, full architecture generation. |
| Multimodal (Accepts Image + Text) | gpt-5.5 | $5.00 | $30.00 | Ultra / Research | Absolute highest tier logic, heavy systems programming, abstract debugging. |
| Text Only (No Image Support) | gpt-4.1-nano | $0.10 | $0.40 | Utility / Parsing | Rapid code linting, syntax formatting, regex building, boilerplate generation. |
| Text Only (No Image Support) | o4-mini | $1.10 | $4.40 | Reasoning Mini | Math-heavy logic, algorithm fixing, edge-case testing, agent loops. |
| Text Only (No Image Support) | o3-mini | $1.10 | $4.40 | Reasoning Mini | Deep logical tree searching, finding obscure multithreading or concurrency bugs. |
| Text Only (No Image Support) | o1 | $15.00 | $60.00 | Frontier Reasoning | Multi-step reasoning problems, planning complex code migrations without human input. |
Tips
The Mini Leverage
Run gpt-4o-mini or gpt-5.4-mini for most coding tasks, then switch to an o-series model when reasoning gets difficult.
Tips
Prompt Caching
Repeated file reads can reduce input costs by 50% thanks to OpenAI prompt caching.