Compare the latest reported prices for GPT, Claude, Gemini, GLM and other models by gateway route and tier.
Text models commonly charge input and output tokens separately, usually in USD per 1 million tokens. Cache reads and cache writes may use different rates. Image generation, video, audio and other multimodal requests can instead be billed per image, second or request. Always check the unit and capability labels shown with the selected model.
Start with model and API compatibility, then compare availability, latency, account tier, context length and priority support. The same model can have different prices and capabilities across routes. Prices on this page are reported by each gateway and refresh dynamically; the final charge is recorded in your request and usage logs.
