Models and pricing

Open Models & Pricing in the Divyam console. Use the current catalog to choose a model and review its rates.

Choose a model

1. Find an available model in the catalog.
2. Read its description and context window.
3. Review input, output, cache-read, and cache-write rates. The cards show rates per million tokens.
4. Select Use this model to prepare its example in Getting Started.
5. Copy the complete model ID, including its prefix, into your application’s configuration.

For example, these guides use divyam-as/gpt-5.6-sol. Use an available ID from your console if the catalog differs.

Keep the requested model consistent

The Getting Started selection changes the generated example. It does not lock an API key or external application to that model. The model value your application sends identifies the requested catalog model and its applicable billing rates.

When changing models, update the application’s request value as well as any connection setting. Check Logs to confirm the requested model for each test. An unknown model ID fails rather than selecting the model shown in Getting Started.

Divyam can route a catalog-model request to a backing model. The completion’s model field identifies the requested catalog model. Review Usage → Requested against routed for recorded routing information.

Compare results and spend

Run the same application test cases for each requested model. Record the model ID, input, expected behavior, actual answer, and request time. Compare the application results with usage and spend for the same period.