Guide to unlimited ai tokens and usage

What are unlimited AI tokens?

Unlimited AI tokens refer to a billing model where you pay only for the text processed, rather than renting access for a fixed period. This approach eliminates monthly subscription fees and ensures you never pay for idle time. You can send as many requests as your rate limits allow, and your account is charged strictly for the input and output tokens used. This model is particularly useful for developers running agents that have variable workloads, as costs scale directly with usage.

Unlike traditional platforms that cap your access or throttle you after a certain number of days, this system allows continuous operation. The term 'unlimited' here describes the absence of hard caps on the number of requests, provided you have sufficient prepaid credit. You control the volume by managing your balance, not by waiting for a monthly reset.

How we calculate token usage

Token usage is calculated based on the exact number of tokens sent in your prompt and the tokens generated in the response. One token typically corresponds to roughly four characters of English text. You can view the token count in the response headers or in the final chunk of a streaming response. This transparency allows you to monitor costs precisely without guessing how much a complex request will cost.

The system counts every token, including system instructions, conversation history, and the final output. If you send a large context window, you pay for all of it. Errors do not consume tokens, and refusals based on content policies are free. This ensures you only pay for successful, usable outputs.

Context window: 100,000 tokens

The model supports a context window of 100,000 tokens, which includes both the prompt and the completion. This large window allows you to pass extensive documentation, codebases, or long conversation histories in a single request. It is ideal for tasks like summarizing large documents or maintaining long-term memory in AI agents.

However, a larger context window does not mean unlimited free usage. You still pay for every token within that window. If you send a 90,000-token prompt, you will be charged for those 90,000 tokens plus the output tokens. The 100,000-token limit is a hard constraint for a single request, ensuring consistent performance and predictable latency.

Max output limits per request

Each request can generate a maximum of 16,000 tokens of output. If you do not specify a max_tokens value, the default limit is 2,048 tokens. This limit applies to the completion generated by the model, not the total context window. You can adjust the max_tokens parameter in your API calls to control the length of the response.

For tasks requiring longer outputs, you may need to implement logic to split the response or continue generation. The API returns a token_usage object in the response, allowing you to track how many tokens were generated. This helps you stay within your budget and plan for large data processing tasks.

Pricing per million tokens

The pricing structure is simple and transparent: $0.15 per 1 million input tokens and $0.50 per 1 million output tokens. You pay only for the tokens you actually use, and errors are free. This makes it cost-effective for high-volume usage, as you only spend when you get a result.

The prepaid model means you must top up your account before you can make requests. Once you have credit, you can use it indefinitely. The credit never expires, so you can top up when prices are favorable or when you have a large project coming up. This flexibility allows you to manage cash flow effectively without worrying about monthly bills.

Crypto top-ups and bonuses

Top-ups are processed exclusively via cryptocurrency, specifically USDT on the TRC20 network or USDC on the Base network. You can top up any whole amount between $20 and $500 per transaction. The system automatically credits your account upon confirmation of the transaction.

Bonus credit is added to your account based on the size of your top-up. Top-ups of $50 or more receive a 5% bonus. Top-ups of $100 or more receive a 10% bonus, and top-ups of $200 or more receive a 15% bonus. This bonus is added to your prepaid balance, effectively lowering your cost per token. There are no limits on how many times you can top up, allowing you to accumulate a large balance for long-term projects.

Rate limits per API key

Each API key is subject to a rate limit of 300 requests per minute and 8 concurrent requests. This ensures fair usage and prevents any single account from monopolizing resources. If you exceed these limits, you will receive a 429 Too Many Requests error. You can implement exponential backoff in your client code to handle these errors gracefully.

The limit of 8 MB per request body prevents excessively large payloads from disrupting the system. If you need higher limits for specific use cases, you can contact support. Each account is limited to one active API key at a time. Generating a new key will replace the old one, so you should update your applications accordingly.

Refunds and credit expiration

Prepaid credit never expires, so you can use it whenever you are ready. If you make a mistake, such as a double charge, you can request a refund through the Support page. The credit is not automatically refunded, so you must initiate the process manually. This policy ensures that your balance remains available for future use, regardless of how much time passes.

The trial credit of $0.50 is valid for 7 days and is non-refundable. It is designed to let you test the API without risking your own funds. After the trial period, the credit expires, and you must top up to continue using the service.