MODELS & PRICING
Full intelligence. A fraction of the price.
Every call reaches the original, full-strength model — never downgraded, never watered down. Official and TokenHey prices side by side, so the savings speak for themselves.
Authentic modelsPay as you goTrackable charges
1 models
| Model | Input / 1M | Cache read / 1M | Output/1M | Discount | context | operate |
|---|---|---|---|---|---|---|
OP GPT-6 Astra OpenAI · openai/gpt-6-astra | $1 $10 | $1 | $5 $50 | -90% | 1.1M |
Token prices are USD / 1M tokens. Image models are billed per request; crossed-out image prices are official per-image reference costs. Cache reads are billed at this rate only when the model response reports a cache hit.
The price is subject to the billing result at the time of actual call. Learn about billing rules →