Yomi API / Integration and billing reference
Yomi API integration and billing: URLs, protocols, models and prices
Find an answer below, then check the model page and documentation for the exact parameters.
References checked:
What is Yomi API, and how do I connect?
Yomi API provides access to multiple models for applications and development tools. To connect, use a Yomi API key, a service URL matching your client protocol, and a model ID you can access. A model provider’s web subscription and API access are different ways to use models.
References: Quick start · API overview and authentication
Which base URL should I use?
Use https://api.yomiapi.com/v1 as the base URL for an OpenAI-compatible SDK, and https://api.yomiapi.com for the Anthropic SDK. For direct HTTP requests, use the complete endpoint for the selected protocol. Keep SDK base URLs distinct from full request endpoints, and avoid adding /v1 twice.
| Client or request method | URL to enter | Usage notes |
|---|---|---|
| OpenAI SDK | https://api.yomiapi.com/v1 | The SDK appends the endpoint path for the selected method. |
| Anthropic SDK | https://api.yomiapi.com | The SDK’s Messages method uses /v1/messages. |
| Gemini native HTTP request | https://api.yomiapi.com/v1beta/models/{model}:generateContent | This is a complete request endpoint. Replace {model} with the model ID. |
References: API overview and authentication · SDK integration
Can every model use the same endpoint?
The endpoint depends on the protocols supported by the model. Check the model detail page’s integration examples, then choose Chat Completions, Responses, Messages, Images or a native Gemini endpoint for your client. Compatibility with a request format does not imply support for every provider API, parameter or capability.
| Protocol | HTTP method and path | Parameter documentation |
|---|---|---|
| Chat Completions | POST /v1/chat/completions | Read the docs |
| Responses | POST /v1/responses | Read the docs |
| Anthropic Messages | POST /v1/messages | Read the docs |
| Images | POST /v1/images/generations | Read the docs |
| Gemini generateContent | POST /v1beta/models/{model}:generateContent | Read the docs |
References: Model Explorer · API overview and authentication
Where do I find a model ID, and what should I check before switching?
Open a model detail page from Model Explorer and copy the exact value shown as “Model ID”. Before switching, check the target model’s protocol, required capabilities, context limits and routes available to your API key. Do not infer the request’s model value or endpoint compatibility from a display name alone.
References: Model Explorer · Model list and queries
How should I compare public prices and actual charges?
Compare prices using the same billing unit, model, route and applicable conditions. Check input, output, cache and tier prices separately; a per-request price is not directly comparable with a price per million tokens. Public model pages show price snapshots. Check Model Explorer for current public prices, and the console for account-specific prices and actual charges.
To estimate costs, measure usage and charges with a representative sample, then scale to your expected request volume. For cache or tiered pricing, also check the applicable conditions.
References: Pricing and billing · Model Explorer
What do availability, time to first token and output speed mean?
Model-page availability uses active probe results from the past 24 hours, while performance metrics come from actual calls. Time to first token (TTFT) measures the wait for the first output; output speed is measured in tokens per second and is not the total request duration. Compare the route, observation window and page snapshot time alongside the metrics.
Historical observations vary with requests, routes and observation windows. They are not a promise of speed or availability for the next request; treat metrics without samples as unknown.
References: Model Explorer
What should I check first when a request fails?
First check the service URL, complete endpoint path, authentication headers, model ID and routes available to your API key, then inspect the HTTP status and error message. Correct parameter or authentication errors before retrying. For rate limits, reduce request frequency and follow the endpoint documentation with bounded backoff retries.
References: API overview and authentication · Quick start
Yomi API