Credits are your account balance.
Credits are not a fixed token package. They work as prepaid balance. When you call a model, TokenSea calculates the cost from the selected model's input and output token prices and deducts it from your balance.
Current models, with live availability shown clearly.
DeepSeek V4 is connected now. The current GPT-5.6 family is listed with planned TokenSea pricing, but is not marked available until an OpenAI or Codex provider is connected. Prices are per 1M tokens in USD.
| Model | Family | Best for | Input / 1M tokens | Output / 1M tokens | Context | Status |
|---|---|---|---|---|---|---|
| deepseek-v4-flash | DeepSeek V4 | Fast general work | $0.14 | $0.28 | 1M | Available |
| deepseek-v4-pro | DeepSeek V4 | Advanced reasoning | $0.435 | $0.87 | 1M | Available |
| gpt-5.6-sol | GPT-5 series | Flagship agents and coding | $0.71 | $4.29 | 1.05M | Provider setup required |
| gpt-5.6-terra | GPT-5 series | Balanced workloads | $0.29 | $1.71 | 1.05M | Provider setup required |
| gpt-5.6-luna | GPT-5 series | High-volume tasks | $0.03 | $0.17 | 1.05M | Provider setup required |
DeepSeek prices show current cache-miss input rates. GPT-5.6 prices are planned TokenSea rates at one seventh of current OpenAI API list prices. Your exact request costs are itemized in your dashboard.
Request cost = input tokens × input price + output tokens × output price
If a request uses 10,000 input tokens and 2,000 output tokens, the final cost depends on the model you selected.
Ready to start building?
Create an API key, add credits in your dashboard, and start calling models through the TokenSea API.
Pricing questions, answered.
Do I need a monthly subscription?
No. TokenSea uses usage-based billing. You only pay for actual model usage.
Are all models priced the same?
No. Each model has its own input and output token pricing.
What are credits?
Credits are prepaid account balance. They are deducted when you make API requests.
Where can I add credits?
Credits are added from Dashboard → Billing.
Can I use one balance across multiple models?
Yes. Your TokenSea balance works across supported models.
Where can I see my usage?
Usage, cost, requests, and remaining balance are available in your dashboard.