Skip to main content
VOID

Void APIHow to buy LLM API access and estimate your costs

How to buy LLM API access and estimate your costs

With Void API, you fund a shared account balance and pay for model usage. An API key authenticates your requests; you do not need to purchase a separate key for each model. Costs depend on the model and the tokens it processes.

Choose a model and compare API pricing

Start with the task: customer responses, document processing, text generation, or coding assistance. Compare models in the public catalog. Prices are listed separately per million input and output tokens, with cache read and write rates where applicable.

An approximate request cost is the number of input tokens multiplied by the input rate, plus the output tokens multiplied by the output rate. For rates quoted per million tokens, divide each token count by one million. Caching and model-specific accounting can affect the final charge.

For example, 2,000 input tokens represent 0.002 million tokens, and 500 output tokens represent 0.0005 million. Multiply those quantities by the current model rates. This is a budget estimate, not a fixed price per request: answer length and actual usage accounting vary.

View current LLM API prices

Add funds to your account balance

Register, sign in, and open the balance funding interface. It shows the payment methods currently available and the final terms of the top-up. Review the payment currency, amount credited, and payment conditions before confirming.

The shared balance funds both API requests and chat usage. A top-up is not an official subscription from a third-party model developer or an unlimited request allowance. Payment, usage charges, and unused-balance refunds are described in the service terms.

Open balance and fundingPayment and refund terms

Create an API key and connect your client

Create a key in the API keys section. Configure your application with the gateway base URL, that key, and an exact model ID from the catalog. Make a short test request before uploading long documents or launching an autonomous coding agent.

Follow the Void API instructions for OpenAI-compatible clients. An inference API key and an account browser session are different credentials. Use the gateway key for model requests, and keep it out of public source code.

Create an API keyMake your first API requestSet up OpenCode, Codex CLI, and other clients

Track costs after your integration is live

Review usage in your account and compare it with your budget. A longer conversation history increases input tokens, while a longer answer increases output tokens. A coding agent may make multiple model calls to complete one task.

If you have a valid promo code, you can use the credited bonus to test the integration. Once that balance is spent, further usage requires funds. Check model availability and current rates before sending requests.

Try the API with promo creditCompare subscriptions and pay-as-you-go API access

Frequently asked questions

Am I buying an API key or model usage?

You add funds to pay for requests to models. Create a key in your account to authenticate those requests; a separate key purchase for each model is not required.

Where can I check prices before paying?

The public catalog shows USD prices per million tokens. The funding interface displays the final top-up terms before you confirm payment.

Does every API request have a fixed price?

No. Charges depend on the selected model, input and output tokens, and applicable cache accounting. Estimate costs with small tests before processing large workloads.