Skip to content

DOCS.md

Latest commit

 

History

History
124 lines (90 loc) · 5.11 KB

DOCS.md

File metadata and controls

124 lines (90 loc) · 5.11 KB

OneLLM.dev API documentation

Welcome to the official OneLLM.dev API documentation! Here you will find:

Table of Contents

It is advised to properly read and understand the documentation before going ahead and making your product. Best of luck!

Authentication

All the API requests require an API Key. Get your API key at OneLLM dashboard after logging in.

Include the key in the Authorization header as a Bearer token:

Authorization: Bearer $APIKEY

An API key is a unique identifier used to authenticate and authorize a user or application when accessing an API (Application Programming Interface). It acts like a password, allowing the API to verify that the request is coming from a legitimate source.

API Endpoint

The BASE_URL for the API is https://onellm.dev/api. This is the primary endpoint for interacting with the language models. It allows you to send a chat conversation and receive a response from the specified model. POST https://onellm.dev/api Note: Streaming responses are not supported. Do not include a stream parameter.

Request Body

All the requests must be a JSON object with the following fields:

Parameter Type Required Description
Model String Yes The model ID to use for the request. See Supported Models for model information.
Messages Array Yes Conversation history as an array of message objects. Each object requires role and content. See an example.
Temperature Number No Controls randomness. 0-2, with 0 being too deterministic and 2 being very random. Default is 1.
Max_tokens Integer No Maximum number of tokens to generate. Server adjusts automatically if it exceeds your balance.
top_p Number No Nucleus sampling probability (0–1). Controls diversity.
stop_sequences Array No Strings that will cause the model to stop generating tokens.

To view more parameters, please go visit the OpenAPI specifications.

Example

Example request:

{
  "model": "GPT-4.1",
  "messages": [
    { "role": "user", "content": "Tell me a joke about computers." }
  ],
  "max_tokens": 50
}

Example response:

{
  "provider": "openai",
  "model": "GPT-4.1",
  "role": "assistant",
  "content": "Why did the computer show up at work late? It had a hard drive!",
  "usage": {
    "input_tokens": 7,
    "output_tokens": 16,
    "total_tokens": 23
  },
  "finish_reason": "stop"
}

Models

Available model IDs (use in model field):

  • GPT Series: GPT-5, GPT-5-Mini, GPT-5-Nano, GPT-5-Chat-Latest, GPT-4.1, GPT-4.1-Mini, GPT-4.1-Nano, GPT-o3, GPT-o3-pro, GPT-o3-DeepResearch, GPT-o3-Mini, GPT-o4-mini, GPT-4o, GPT-4o-mini, GPT-o1, GPT-o1-Mini

  • Anthropic (Claude-like): Opus-4, Sonnet-4, Haiku-3.5, Opus-3, Sonnet-3.7, Haiku-3

  • DeepSeek: DeepSeek-Reasoner, DeepSeek-Chat

  • Flash/Preview: 2.5-Flash-preview, 2.5-Pro-preview, 2.0-Flash, 2.0-Flash-lite, 1.5-Flash, 1.5-Flash-8B, 1.5-Pro

  • Mistral & Variants: Mistral-Medium-3, Magistral-Medium, Codestral, Devstral-Medium, Mistral-Saba, Mistral-Large, Pixtral-Large, Ministral-8B-24.10, Ministral-3B-24.10, Mistral-Small-3.2, Magistral-Small, Devstral-Small, Pixtral-12B, Mistral-NeMo, Mistral-7B, Mixtral-8x7B, Mixtral-8x22B

Error Handling

The API may return different error codes. Common ones are:

Status Code Meaning Reason
400 Bad Request Invalid JSON or missing required fields.
401 Unauthorized API key missing or invalid.
403 Forbidden You don't have access to the requested resource
404 Not Found Invalid Endpoint.
405 Method not allowed Endpoint does not support the HTTP method used.
422 Unprocessable Entity Invalid type or format in the request body.
429 Too Many Requests Rate limit exceeded. Try again later.
500 Internal Server Error Something went wrong on the server side.

Please be sure to check the response body for extra error detail. If you encounter a error you cannot debug, please be sure to join our Discord Support server.

Important Notes

  • Streaming is not supported. The model sends complete response for now. But we do have someone working on it, so it won't be long before this feature is implemented.

  • max_tokens will auto-adjust if it’s larger than your allowed balance.

  • A minimum balance of $0.10 is required to make API requests.

  • The response includes usage info:

    • input_tokens: number of tokens sent in the request.

    • output_tokens: number of tokens generated by the model.

    • total_tokens: combined count of both.