OpenAI
GPT-5.4 with maximum reasoning effort — for the hardest multi-step problems where you want the model to spend most of its budget thinking before answering.
OpenAI
GPT-5.4 with high reasoning effort — for complex debugging, cross-file analysis, and design tradeoffs.
OpenAI
GPT-5.4 with medium reasoning effort — for multi-step refactors and design choices that benefit from a few extra thinking tokens.
OpenAI
GPT-5.4 with low reasoning effort — light reasoning for day-to-day coding tasks; faster and cheaper than higher tiers.
OpenAI
Cheapest OpenAI gpt-image-2 channel. Sync image-to-image edit API with multi-image fusion (up to 10), 1K/2K/4K output and quality control. Flat $0.03 per image.
OpenAI
OpenAI GPT Image 2 (beta channel). Text-to-image and multi-image editing (up to 16 reference images), aspect-ratio controlled output. Independent channel from the primary gpt-image-2 route for redundancy. Flat $0.03 per image.
OpenAI
OpenAI gpt-image-2. Text-to-image and multi-image editing (up to 10 reference images), aspect-ratio control, flat $0.03 per image at Medium/High quality.
OpenAI
Uses more compute to think deeper and deliver consistently better answers. Supports multi-turn model interactions and advanced API features.
OpenAI
Our frontier model for complex professional work.
OpenAI
Powerful model with excellent performance and efficiency.
OpenAI
Snapshot version of GPT-5.1 for reproducible outputs.
OpenAI
GPT-5 with integrated web search for real-time information.
OpenAI
Snapshot version of GPT-5 Search API for stable deployments.
OpenAI
Versatile model for general-purpose tasks.
OpenAI
Fast reasoning model with efficient output generation.
OpenAI
Advanced reasoning model for complex analytical tasks.
OpenAI
Small embedding model, efficient and cost-effective for most use cases.
OpenAI
Large embedding model for higher accuracy and flexible dimensions.
