Which model runs my task?

Use for: Quick chat, intake, classification, light planning.

Fastest and cheapest. Used by the New Task chat and short tool calls.

Use for: Multi-step workflows, long-context research, multimodal analysis.

Best price-per-token for big context and image-text reasoning.

Use for: Tasks where nuance and tone matter — writing, sensitive emails, legal-flavoured drafts.

Strongest reasoning and writing in our gateway. More expensive — used selectively.

Use for: High-volume drafting where you want GPT-quality at a fraction of the cost.

Keeps most of GPT-5's strengths at roughly a quarter of the price.

Use for: Generating screenshots, charts, social images on the fly.

Fast image generation built into the same gateway — no separate vendor.