> ## Documentation Index
> Fetch the complete documentation index at: https://docs.codeflare.cc/llms.txt
> Use this file to discover all available pages before exploring further.

# Models and protocols

> Copy the platform model ID and check supported protocols and capabilities.

export const ConsoleLink = ({path = "/console", children}) => <a href={"http://localhost:3001" + path} target="_blank" rel="noopener noreferrer">{children}</a>;

## Copy the model ID

Open the <ConsoleLink path="/console/models">model catalog ↗</ConsoleLink>, select a model, and use **Copy model ID** beside its details title. The request's `model` is the platform model ID, not its display name or an upstream name you guessed.

| Usage | Required protocol |
| - | - |
| Responses, Codex CLI | `responses` |
| Chat Completions | `chat-completions` |
| Anthropic Messages, Claude Code | `anthropic-messages` |
| Native Gemini API | `gemini` |
| TypeSafe structured judgments | `systemone` |

The OpenAI SDK has both Responses and Chat Completions methods. Check the selected model's declared protocols before choosing a method; some models expose multiple protocols.

## Capabilities and prices

Details show context size, output limits, capabilities, and current prices. Multimodal input, tools, and background execution also depend on channels, routes, and upstream implementation. Protocol compatibility does not imply support for every provider feature.

Requests depend on balance, key permissions, model enablement, and route health. An empty catalog requires administrator configuration.

<Note>This documentation does not maintain a live model catalog or price list. Use the console as the source of truth; see [Capabilities and limits](/en/capabilities).</Note>


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.