LLM Text Generation

Use this section when you want to integrate OmniRouters for chat, text generation, reasoning, or protocol migration.

OmniRouters supports multiple request protocols for LLM text generation. The most important rule is simple:

All models enabled in your OmniRouters account can be called through the OpenAI-compatible protocol.

That means /v1/chat/completions and /v1/responses are the safest default choices when you want one request style that works across providers and model families.

Recommended Default

For most new integrations, start with one of these two OpenAI-compatible routes:

• /v1/chat/completions for broad SDK compatibility and simple chat workflows
• /v1/responses for newer structured workflows, tool use, and response formatting

Use the Claude or Gemini protocol pages when you are migrating an existing client, SDK, or payload format and want to minimize request-shape changes.

Available Protocols

Protocol	OmniRouters endpoint	Best for	Official docs
OpenAI Chat Completions	POST /v1/chat/completions	Broad compatibility, simple chat, fastest migration from existing OpenAI chat clients	OpenAI Chat API reference
OpenAI Responses	POST /v1/responses	Newer structured workflows, tool calling, reasoning-oriented integrations	OpenAI Responses API reference
Claude Messages	POST /v1/messages	Anthropic / Claude SDK migration with a Claude-style payload	Anthropic Messages API reference
Gemini Generate Content	POST /v1beta/models/{model}:generateContent	Google Gemini SDK migration with contents / parts request format	Google Gemini generateContent reference

OmniRouters Conventions

Before you choose a protocol, keep these OmniRouters-specific rules in mind:

• Base URL: https://omnirouters.com
• OpenAI-compatible routes usually live under https://omnirouters.com/v1
• Gemini-style routes use https://omnirouters.com/v1beta/...
• Authentication uses Authorization: Bearer <your-api-key> on OmniRouters, even when the upstream provider uses another auth header
• Model names must match the models enabled in your OmniRouters account
• Use https://omnirouters.com/keys to create and manage API keys

Which Protocol Should You Pick?

If you are building a new app

Start with the OpenAI-compatible protocol.

• Choose OpenAI Chat Completions if you want the most common request shape
• Choose OpenAI Responses if your app needs structured outputs or tool workflows

If you already use Claude

Use the Claude Messages page first. It keeps the request body closer to Anthropic's native format.

If you already use Gemini

Use the Gemini Generate Content page first. It keeps the contents / parts structure familiar.

If you want the shortest path with the widest compatibility

Start with OpenAI Chat Completions. On OmniRouters, this is the cleanest default for cross-provider model access.

Continue Reading

1. Read OpenAI Chat Completions
2. Review Protocol Comparison
3. Open the full Apifox API Reference