Documentation

Setup Guide

Get Omega Plus running inside your IDE or AI client with the real proxy URL, Omega model IDs, and Anthropic-compatible endpoints.

Drop-in Compatible Per-Key Budgets Proxy Auth 12 Models

Prerequisites

Node.js 18 or newer

Required for running the Omega Plus npm setup wizard with npx.

Omega Plus API key

Create or receive a customer API key before connecting a coding tool.

Supported coding tool

Claude Code, Codex, OpenCode, Cline, Roo Code, Kilo Code, Factory Droid, OpenClaw, Zed, Warp, Continue, Aider, Cursor, Windsurf, or any OpenAI/Anthropic-compatible client.

If Node.js is not installed, download the official installer from nodejs.org, then reopen your terminal and run the setup command.

Quick Setup

Run the Omega Plus setup wizard, paste your API key, then choose the coding tool you want to configure.

Main Command

npx -y @kesarcloud/omega-plus-cli@latest

Connection Values

API origin: https://api.omegaplusapi.com
Anthropic-compatible URL: https://api.omegaplusapi.com/v1
OpenAI-compatible URL: https://api.omegaplusapi.com/api/v1
API Key: YOUR_API_KEY
Primary model: omega-plus
Fast model: omega-plus

Interactive Setup Flow

The wizard keeps the API key out of the shell history, detects installed tools, and writes the right config for the selected tool.

Omega Plus banner appears in the terminal.

User pastes the Omega API key when prompted.

Wizard validates the key and lists available coding tools.

User enters the tool number, then Omega Plus writes the config.

User opens or restarts the selected coding tool.

Claude Code CLI

Recommended: run the Omega Plus wizard and select Claude Code from the tool list. The wizard writes ~/.claude/settings.json for you.

Recommended Setup

npx -y @kesarcloud/omega-plus-cli@latest

Optional Direct Command

npx -y @kesarcloud/omega-plus-cli@latest configure --tool claude --api-key YOUR_API_KEY

Manual fallback: create or edit this file, then restart Claude Code.

~/.claude/settings.json

{
  "env": {
    "ANTHROPIC_AUTH_TOKEN": "YOUR_API_KEY",
    "ANTHROPIC_BASE_URL": "https://api.omegaplusapi.com/v1",
    "ANTHROPIC_MODEL": "omega-plus",
    "ANTHROPIC_SMALL_FAST_MODEL": "omega-plus",
    "ANTHROPIC_DEFAULT_OPUS_MODEL": "claude-opus-4-8",
    "ANTHROPIC_DEFAULT_SONNET_MODEL": "claude-sonnet-4-6",
    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "omega-plus",
    "CLAUDE_CODE_SUBAGENT_MODEL": "omega-plus",
    "CLAUDE_CODE_ENABLE_GATEWAY_MODEL_DISCOVERY": "1",
    "CLAUDE_CODE_MAX_OUTPUT_TOKENS": "128000",
    "CLAUDE_CODE_DISABLE_NONESSENTIAL_TRAFFIC": "1"
  },
  "model": "omega-plus",
  "hasCompletedOnboarding": true
}

OpenAI Codex CLI

Recommended: run the Omega Plus wizard and select OpenAI Codex CLI. Codex uses the OpenAI-compatible Omega endpoint.

Recommended Setup

npx -y @kesarcloud/omega-plus-cli@latest

Optional Direct Command

npx -y @kesarcloud/omega-plus-cli@latest configure --tool codex --api-key YOUR_API_KEY

Manual fallback: set the Omega provider in ~/.codex/config.toml, then keep the API key in ~/.codex/auth.json.

~/.codex/config.toml

model = "omega-plus"
model_provider = "omega"

[model_providers.omega]
name = "Omega Plus"
base_url = "https://api.omegaplusapi.com/api/v1"
wire_api = "responses"
env_key = "OPENAI_API_KEY"

~/.codex/auth.json

{
  "OPENAI_API_KEY": "YOUR_API_KEY"
}

Hermes Agent

Hermes Agent is a self-improving autonomous agent by Nous Research with a built-in learning loop, persistent memory, and skills system. It speaks the OpenAI API, so pointing it at the Omega Plus endpoint takes one config block.

Manual Setup

# ~/.hermes/config.yaml
model:
  default: "omega-plus"
  provider: "custom"
  base_url: "https://api.omegaplusapi.com/api/v1"
  api_key: "YOUR_API_KEY"

Manual fallback: edit ~/.hermes/config.yaml and point the custom provider at the Omega OpenAI-compatible base URL.

~/.hermes/config.yaml

model:
  default: "omega-plus"        # pick any of the 12 Omega Plus available models
  provider: "custom"           # OpenAI-compatible endpoint
  base_url: "https://api.omegaplusapi.com/api/v1"
  api_key: "YOUR_API_KEY"      # your Omega Plus API key (oc_...)

Alternative: save the key to ~/.hermes/.env

echo "OPENAI_API_KEY=YOUR_API_KEY" >> ~/.hermes/.env

Inside a running Hermes session, switch models with /model omega-plus, /model claude-opus-4-8, or any of the 12 available model IDs.

Install Hermes Agent

curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash

After install, run hermes model and pick Custom endpoint from the menu — paste the OpenAI base URL and your API key when prompted.

Reference: Custom & Self-Hosted LLM Providers in Hermes docs.

OpenCode

Recommended: run the Omega Plus wizard and select OpenCode. The wizard writes the provider block in ~/.config/opencode/opencode.json.

Recommended Setup

npx -y @kesarcloud/omega-plus-cli@latest

Optional Direct Command

npx -y @kesarcloud/omega-plus-cli@latest configure --tool opencode --api-key YOUR_API_KEY

Manual fallback: add the Omega Plus provider and choose the Omega model.

~/.config/opencode/opencode.json

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "omegaplusapi": {
      "npm": "@ai-sdk/anthropic",
      "name": "Omega Plus",
      "options": {
        "baseURL": "https://api.omegaplusapi.com/v1",
        "apiKey": "YOUR_API_KEY"
      },
      "models": {
        "omega-plus": { "name": "Omega Plus" },
        "claude-opus-4-6": { "name": "Claude Opus 4.6" },
        "claude-opus-4-7": { "name": "Claude Opus 4.7" },
        "claude-opus-4-8": { "name": "Claude Opus 4.8" },
        "gpt-5.5": { "name": "GPT 5.5" },
        "claude-sonnet-4-6": { "name": "Claude Sonnet 4.6" },
        "gemini-3.1-pro": { "name": "Gemini 3.1 Pro" },
        "deepseek-v4-pro": { "name": "DeepSeek V4 Pro" },
        "qwen-3.7-plus": { "name": "Qwen 3.7 Plus" },
        "minimax-m3": { "name": "MiniMax M3" },
        "glm-5.2": { "name": "GLM 5.2" },
        "claude-fable-5": { "name": "Fable 5" }
      }
    }
  }
}

Cline

Recommended: run the Omega Plus wizard and select Cline. If Cline is not installed, install the extension first, then rerun the wizard.

Recommended Setup

npx -y @kesarcloud/omega-plus-cli@latest

Optional Direct Command

npx -y @kesarcloud/omega-plus-cli@latest configure --tool cline --api-key YOUR_API_KEY

Manual fallback: use Anthropic-compatible mode with the Omega endpoint.

Cline Settings

{
  "actModeApiProvider": "anthropic",
  "planModeApiProvider": "anthropic",
  "anthropicBaseUrl": "https://api.omegaplusapi.com/v1",
  "anthropicModelId": "omega-plus",
  "planModeAnthropicModelId": "omega-plus"
}

Kilo Code

Recommended: run the Omega Plus wizard and select Kilo Code. The wizard configures Kilo with the Anthropic-compatible Omega endpoint.

Recommended Setup

npx -y @kesarcloud/omega-plus-cli@latest

Optional Direct Command

npx -y @kesarcloud/omega-plus-cli@latest configure --tool kilo --api-key YOUR_API_KEY

Manual fallback: add the Anthropic auth block for Omega Plus.

Kilo Auth

{
  "anthropic": {
    "type": "api-key",
    "apiKey": "YOUR_API_KEY",
    "baseUrl": "https://api.omegaplusapi.com/v1",
    "model": "omega-plus"
  }
}

Factory Droid

Recommended: run the Omega Plus wizard and select Factory Droid. If Droid is missing, install it first, then rerun the wizard.

Recommended Setup

npx -y @kesarcloud/omega-plus-cli@latest

Optional Direct Command

npx -y @kesarcloud/omega-plus-cli@latest configure --tool droid --api-key YOUR_API_KEY

Manual fallback: add a custom Anthropic-compatible model entry.

Droid Custom Model

{
  "model": "omega-plus",
  "id": "custom:OmegaPlus-0",
  "baseUrl": "https://api.omegaplusapi.com/v1",
  "apiKey": "YOUR_API_KEY",
  "displayName": "Omega Plus",
  "provider": "anthropic"
}

Roo Code

Roo Code is a VS Code extension that supports OpenAI-compatible and Anthropic-compatible custom providers. Configure it manually using the Omega Anthropic endpoint.

Step 1

Install the Roo Code extension from the VS Code Marketplace.

Step 2

Open Roo Code settings (gear icon in the Roo Code panel).

Step 3

Set API Provider to Anthropic.

Step 4

Set Base URL to the Omega Anthropic URL below.

Step 5

Paste your Omega API key.

Step 6

Set Model to omega-plus or any of the 12 available model IDs.

Step 7

Save and restart VS Code.

Roo Code settings.json

{
  "roo-cline.apiProvider": "anthropic",
  "roo-cline.anthropicBaseUrl": "https://api.omegaplusapi.com/v1",
  "roo-cline.apiKey": "YOUR_API_KEY",
  "roo-cline.model": "omega-plus"
}

Reference: Roo Code Anthropic Provider Docs

Zed IDE

Zed is a fast IDE with first-class OpenAI-compatible provider support. Configure it via settings JSON or the UI.

Step 1

Install Zed from zed.dev/download.

Step 2

Open settings with cmd+, (macOS) or ctrl+, (Linux).

Step 3

Under LLM Providers, click Add Provider.

Step 4

Choose OpenAI Compatible and enter a provider name (e.g. Omega Plus).

Step 5

Set the Base URL to the Omega OpenAI URL below.

Step 6

Paste your Omega API key.

Step 7

Add available models (e.g. omega-plus, claude-opus-4-8).

Step 8

Save and select the model from Zed's model picker.

Zed settings.json (openai_compatible)

{
  "language_models": {
    "openai_compatible": {
      "omega_plus": {
        "api_url": "https://api.omegaplusapi.com/api/v1",
        "api_key": "YOUR_API_KEY",
        "available_models": [
          { "name": "omega-plus", "display_name": "Omega Plus" },
          { "name": "claude-opus-4-8", "display_name": "Claude Opus 4.8" },
          { "name": "gpt-5.5", "display_name": "GPT 5.5" },
          { "name": "minimax-m3", "display_name": "MiniMax M3" },
          { "name": "glm-5.2", "display_name": "GLM 5.2" },
          { "name": "claude-fable-5", "display_name": "Fable 5" }
        ]
      }
    }
  }
}

Reference: Zed AI API Access Docs

Warp Terminal

Warp is an AI-powered terminal that supports custom OpenAI-compatible inference endpoints.

Note: Warp requires the custom endpoint to be publicly reachable (HTTPS). localhost URLs are not accepted. Use the full Omega Plus domain URL.

Step 1

Install Warp from warp.dev/download.

Step 2

Open Warp Settings (gear icon or cmd+,).

Step 3

Search for inference endpoint or navigate to AI Settings.

Step 4

Add a custom inference endpoint with the Omega OpenAI URL.

Step 5

Paste your Omega API key.

Step 6

Select a model (e.g. omega-plus).

Step 7

Save and test with a prompt in Warp.

Warp Custom Endpoint

Endpoint URL: https://api.omegaplusapi.com/api/v1
API Key: YOUR_API_KEY
Model: omega-plus

Reference: Warp Custom Inference Docs

OpenClaw

OpenClaw is an AI assistant with config-first custom provider support for both OpenAI-compatible and Anthropic-compatible APIs.

Step 1

Install OpenClaw from docs.openclaw.ai.

Step 2

Open or create ~/.openclaw/openclaw.json.

Step 3

Add the Omega Plus provider block below.

Step 4

Save and restart OpenClaw.

~/.openclaw/openclaw.json

{
  "models": {
    "providers": {
      "omega": {
        "api": "openai-completions",
        "baseUrl": "https://api.omegaplusapi.com/api/v1",
        "apiKey": "YOUR_API_KEY",
        "models": [
          { "id": "omega-plus", "name": "Omega Plus" },
          { "id": "claude-opus-4-8", "name": "Claude Opus 4.8" },
          { "id": "gpt-5.5", "name": "GPT 5.5" },
          { "id": "minimax-m3", "name": "MiniMax M3" },
          { "id": "glm-5.2", "name": "GLM 5.2" },
          { "id": "claude-fable-5", "name": "Fable 5" }
        ]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "omega-plus"
      }
    }
  }
}

Reference: OpenClaw Config Docs

Continue

Continue is an open-source AI coding assistant for VS Code and JetBrains. It supports custom OpenAI-compatible providers via config.

Step 1

Install the Continue extension from the VS Code Marketplace.

Step 2

Open Continue config (gear icon → Settings or ~/.continue/config.yaml).

Step 3

Add the Omega Plus model block below.

Step 4

Save and reload VS Code.

~/.continue/config.yaml

models:
  - name: Omega Plus
    provider: openai
    apiBase: https://api.omegaplusapi.com/api/v1
    apiKey: YOUR_API_KEY
    model: omega-plus
  - name: Claude Opus 4.8
    provider: openai
    apiBase: https://api.omegaplusapi.com/api/v1
    apiKey: YOUR_API_KEY
    model: claude-opus-4-8
  - name: MiniMax M3
    provider: openai
    apiBase: https://api.omegaplusapi.com/api/v1
    apiKey: YOUR_API_KEY
    model: minimax-m3
  - name: GLM 5.2
    provider: openai
    apiBase: https://api.omegaplusapi.com/api/v1
    apiKey: YOUR_API_KEY
    model: glm-5.2

Reference: Continue OpenAI Provider Docs

Aider

Aider is a CLI-based AI pair programmer. It supports any OpenAI-compatible endpoint via environment variables.

Step 1

Install Aider: pip install aider-chat

Step 2

Set the environment variables below in your shell or .env file.

Step 3

Run Aider: aider --model openai/omega-plus

Environment Variables

export OPENAI_API_BASE=https://api.omegaplusapi.com/api/v1
export OPENAI_API_KEY=YOUR_API_KEY

# Then run:
aider --model openai/omega-plus

# Or use any of the 12 models:
aider --model openai/claude-opus-4-8
aider --model openai/gpt-5.5
aider --model openai/minimax-m3
aider --model openai/glm-5.2
aider --model openai/claude-fable-5

Reference: Aider OpenAI-Compatible Docs

Cursor

Cursor is an AI-first code editor. It supports BYOK for named providers (OpenAI, Anthropic) with an optional custom Base URL. There is no generic custom OpenAI-compatible provider flow.

Limitation: Cursor does not support a generic OpenAI-compatible endpoint. You can set the OpenAI API Key with a custom Base URL, but the model picker is limited to Cursor's built-in model list. For full Omega Plus model access, use Claude Code, OpenCode, or Cline instead.

Step 1

Open Cursor Settings → Models.

Step 2

Enable OpenAI API Key.

Step 3

Set the Base URL to the Omega OpenAI URL.

Step 4

Paste your Omega API key as the OpenAI API Key.

Step 5

Click Verify to test the connection.

Cursor Settings

OpenAI API Key: YOUR_API_KEY
Base URL: https://api.omegaplusapi.com/api/v1
Model: (select from Cursor's model picker)

Reference: Cursor API Keys Docs

Windsurf

Windsurf is an AI-first IDE by Codeium. Public docs do not describe a generic custom OpenAI-compatible provider flow.

Limitation: Windsurf's official docs focus on built-in models and named-provider BYOK. A generic custom endpoint flow is not clearly documented. Try the Anthropic provider with the Omega Base URL, or use Claude Code / OpenCode / Cline for full Omega Plus model access.

Step 1

Open Windsurf AI Settings.

Step 2

Look for Add custom provider or Anthropic provider settings.

Step 3

If available, set Base URL to the Omega Anthropic URL and paste your API key.

Windsurf Settings (if custom provider is available)

Base URL: https://api.omegaplusapi.com/v1
API Key: YOUR_API_KEY
Model: omega-plus

Reference: Windsurf Models Docs

Manual Configuration Reference

Use this section for tools that are not automated by the Omega Plus npm wizard, or when you prefer to configure a client manually.

Generic OpenAI-Compatible Client

OPENAI_API_KEY=YOUR_API_KEY
OPENAI_BASE_URL=https://api.omegaplusapi.com/api/v1
MODEL=omega-plus

Generic Anthropic-Compatible Client

ANTHROPIC_AUTH_TOKEN=YOUR_API_KEY
ANTHROPIC_BASE_URL=https://api.omegaplusapi.com/v1
MODEL=omega-plus

Authentication

Proxy endpoints accept API key authentication through either header. Dashboard endpoints use JWT tokens returned from the login endpoint.

Header Options

x-api-key: YOUR_API_KEY
# OR
Authorization: Bearer YOUR_API_KEY

POST/v1/messagesAPI Key

Messages

Create a message with Anthropic-compatible request and response shape. Set stream: true for SSE streaming.

Request

{
  "model": "omega-plus",
  "max_tokens": 1024,
  "messages": [
    { "role": "user", "content": "Hello, Omega!" }
  ],
  "stream": false
}

Image & Vision

All 12 Omega models support image input. Send images via /v1/messages (Anthropic format) or /api/v1/chat/completions (OpenAI format).

Anthropic Format (POST /v1/messages)

Request

{
  "model": "claude-fable-5",
  "max_tokens": 256,
  "messages": [{
    "role": "user",
    "content": [
      {
        "type": "image",
        "source": {
          "type": "base64",
          "media_type": "image/png",
          "data": "<base64-encoded-image>"
        }
      },
      {
        "type": "text",
        "text": "What do you see in this image?"
      }
    ]
  }]
}

OpenAI Format (POST /api/v1/chat/completions)

Request

{
  "model": "claude-fable-5",
  "max_tokens": 256,
  "messages": [{
    "role": "user",
    "content": [
      {
        "type": "image_url",
        "image_url": {
          "url": "data:image/png;base64,<base64-encoded-image>"
        }
      },
      {
        "type": "text",
        "text": "What do you see in this image?"
      }
    ]
  }]
}

Vision Capability

Model	Vision	Method
Omega Plus	Yes	Native
Claude Opus 4.6 / 4.7 / 4.8	Yes	Native
Claude Sonnet 4.6	Yes	Native
Fable 5	Yes	Native
GPT 5.5	Yes	Native
Gemini 3.1 Pro	Yes	Native
MiniMax M3	Yes	Native (multimodal)
DeepSeek V4 Pro	Bridge	Omega Vision Bridge
Qwen 3.7 Plus	Bridge	Omega Vision Bridge
GLM 5.2	Bridge	Omega Vision Bridge

Omega Vision Bridge —DeepSeek V4 Pro, Qwen 3.7 Plus, and GLM 5.2 don't natively support vision. Omega Vision Bridge automatically describes images to text, then sends the text description to the model. Fully transparent — no configuration needed, images work out of the box.

GET/v1/modelsNone

Models

List all available public Omega model IDs.

Response

{
  "data": [
    {
      "id": "omega-plus",
      "object": "model",
      "display_name": "Omega Plus"
    },
    {
      "id": "claude-opus-4-6",
      "object": "model",
      "display_name": "Claude Opus 4.6"
    },
    {
      "id": "claude-opus-4-7",
      "object": "model",
      "display_name": "Claude Opus 4.7"
    },
    {
      "id": "claude-opus-4-8",
      "object": "model",
      "display_name": "Claude Opus 4.8"
    },
    {
      "id": "gpt-5.5",
      "object": "model",
      "display_name": "GPT 5.5"
    },
    {
      "id": "claude-sonnet-4-6",
      "object": "model",
      "display_name": "Claude Sonnet 4.6"
    },
    {
      "id": "gemini-3.1-pro",
      "object": "model",
      "display_name": "Gemini 3.1 Pro"
    },
    {
      "id": "deepseek-v4-pro",
      "object": "model",
      "display_name": "DeepSeek V4 Pro"
    },
    {
      "id": "qwen-3.7-plus",
      "object": "model",
      "display_name": "Qwen 3.7 Plus"
    },
    {
      "id": "minimax-m3",
      "object": "model",
      "display_name": "MiniMax M3"
    },
    {
      "id": "glm-5.2",
      "object": "model",
      "display_name": "GLM 5.2"
    },
    {
      "id": "claude-fable-5",
      "object": "model",
      "display_name": "Fable 5"
    }
  ]
}

POST/v1/messages/count_tokensAPI Key

Token Counting

Count tokens for a message without sending it to a provider.

Request

{
  "model": "omega-plus",
  "messages": [
    { "role": "user", "content": "How many tokens is this?" }
  ]
}

POST/api/auth/loginEmail + Password

Dashboard Auth

Use this route to sign in to dashboard sessions.

Request

{
  "email": "YOUR_EMAIL",
  "password": "YOUR_PASSWORD"
}

Available Models

High ReasoningNative Vision

Omega Plus

omega-plus

Frontier coding proxy pool — clean, error-free, production-grade output.

High ReasoningNative Vision

Claude Opus 4.6

claude-opus-4-6

Anthropic Claude Opus 4.6 — high reasoning for complex analysis.

Max ReasoningNative Vision

Claude Opus 4.7

claude-opus-4-7

Anthropic Claude Opus 4.7 — maximum reasoning for frontier tasks.

High ReasoningNative Vision

Claude Opus 4.8

claude-opus-4-8

Anthropic Claude Opus 4.8 — latest Opus with 1M context.

Max ReasoningNative Vision

GPT 5.5

gpt-5.5

OpenAI GPT 5.5 — maximum reasoning for advanced AI tasks.

High ReasoningNative Vision

Claude Sonnet 4.6

claude-sonnet-4-6

Anthropic Claude Sonnet 4.6 — 1M context, high reasoning.

High ReasoningNative Vision

Gemini 3.1 Pro

gemini-3.1-pro

Google Gemini 3.1 Pro — 1M context, full vision support.

Medium ReasoningVision via Bridge

DeepSeek V4 Pro

deepseek-v4-pro

DeepSeek V4 Pro — 1.6T params, MoE, 1M context. Vision via Omega Vision Bridge.

Medium ReasoningVision via Bridge

Qwen 3.7 Plus

qwen-3.7-plus

Qwen 3.7 Plus — 1M context, 128K output. Vision via Omega Vision Bridge.

High ReasoningNative Vision

MiniMax M3

minimax-m3

MiniMax M3 — 428B MoE, MSA architecture, 1M context, multimodal.

High ReasoningVision via Bridge

GLM 5.2

glm-5.2

GLM 5.2 by Z.ai (Zhipu AI) — 753B MoE, 1M context, 131K output. Vision via Omega Vision Bridge.

High ReasoningNative Vision

Fable 5

claude-fable-5

Anthropic Fable 5 — most capable model, 1M context, adaptive thinking, 128K output.

Build Notes

Built-in search and vision

Omega Plus does not require MCP servers or third-party tools for web search and image understanding. All 12 models include built-in web search capability.

Vision capability

All 12 Omega models support image input. Native Vision: Anthropic (Opus, Sonnet, Fable 5), OpenAI (GPT 5.5), Google (Gemini 3.1 Pro), MiniMax (M3), and SingleByte Infotech Pvt Ltd (Omega Plus) models have built-in vision. Omega Vision Bridge:DeepSeek V4 Pro, Qwen 3.7 Plus, and GLM 5.2 don't natively support vision — Omega Vision Bridge automatically describes images to text, then sends the text description to the model. Fully transparent, no configuration needed.

Troubleshooting

Connection errors

Confirm the API server is running at https://api.omegaplusapi.com.

Model not found

Use exact model IDs from the Available Models section.

Rate limited

Your five-hour token window or RPM cap may be exhausted.

Changes not applying

Restart your IDE after any config change.

Cursor/Windsurf not routing

Make sure the base URL ends with /v1.