OpenAI
GPT-4o, GPT-4o-mini...
RABAH.AI
Multiply human capability, time, and freedom
One gateway to leading models. Choose freely, route intelligently, and stay private.
Balance reliability, cost, and capability without lock-in.
Choose the right model. Never be locked into one.
Use one RABAH.AI API to access the latest models from OpenAI, Anthropic, Google, xAI, and more. Integrate once, keep your options open, and move as the frontier changes.
Anthropic
Claude 3.5 Sonnet...
Gemini 1.5 Pro...
xAI
Grok...
More Models
Continuously onboarding...
One gateway. Your choice.
A standardized, OpenAI-compatible API connects vendors and models. Change one base_url and keep control of where your application goes next.
Put every request on the right path
Route across model capability, cost, latency, and availability to find a better fit for each task and constraint.
Reliability and privacy by design
Built around dependable access, streaming responses, and clear data boundaries. Request and response bodies are not stored by default.
See where every token goes
Track requests, usage, and token consumption in one place. Understand cost and runtime behavior through real data.
Keep the Agent on your computer. Carry the controls anywhere.
Desktop Companion keeps your local Agent online and configures the API in one click. The mobile app lets you chat, review progress, handle approvals, and take remote control.
Step 1 · Desktop
Bring your computer Agent online
Install RABAH Companion, sign in with the same account, and enable remote access. Select an API key when needed and apply it to Codex and the remote Agent in one click.
Download for Mac macOS 14+ · Apple silicon For this device Available Download for Windows Windows 10/11 · x64 For this device Available
- Always-on service
- Remote Agent
- One-click API setup
Encrypted link Step 2 · Mobile
Connect from your phone
Use RABAH Chat on its own, then connect to an online computer to send tasks, view sessions, handle approvals, stop, or continue your Agent.
iPhone and iPad App Store For this device Available on TestFlight Android Google Play For this device Coming soon
Sign in with the same RABAH account and your online computers appear automatically.
Sign in with the same RABAH account and your online computers appear automatically. 
Runs on your computer. For remote execution, your computer must remain online and awake. The RABAH mobile app never moves your local Agent into the cloud.
Stay free from lock-in
Keep the OpenAI SDK and request format you already use. Switch with one base_url change and no application rewrite.
Minutes, not weeks
Get an API key, point to the gateway, and run your first streaming request while your time stays focused on the product.
Bodies private by default
Request and response bodies are not persisted by default. Only token usage is recorded for billing, reducing unnecessary data exposure.
End-to-end encrypted chat
Built-in E2EE keeps sensitive conversations private between sending and receiving endpoints.
You create.
RABAH.AI carries the complexity.
Call models from different vendors through familiar interfaces. One infrastructure layer absorbs the differences in compatibility, routing, usage, and cost.
Run Your First Request POST
REQUEST https://api.rabah.ai/v1/chat/completions 1curl -X POST "https://api.rabah.ai/v1/chat/completions" \
2-H "Authorization: Bearer sk-••••" \
3-d '{
4"model": "your-model",
5"messages": [
6 { "role": "user", "content": "..." }
7]
8}' POST
REQUEST https://api.rabah.ai/v1/responses 1curl -X POST "https://api.rabah.ai/v1/responses" \
2-H "Authorization: Bearer sk-••••" \
3-d '{
4"model": "your-model",
5"input": "..."
6}' POST
REQUEST https://api.rabah.ai/v1/messages 1curl -X POST "https://api.rabah.ai/v1/messages" \
2-H "x-api-key: sk-••••" \
3-H "anthropic-version: 2023-06-01" \
4-d '{
5"model": "your-model",
6"max_tokens": 1024,
7"messages": [
8 { "role": "user", "content": "..." }
9]
10}' POST
REQUEST https://api.rabah.ai/v1beta/models/{model}:generateContent 1curl -X POST "https://api.rabah.ai/v1beta/models/{model}:generateContent" \
2-H "x-goog-api-key: sk-••••" \
3-d '{
4"contents": [
5 { "role": "user",
6 "parts": [{ "text": "..." }] }
7]
8}'What We Believe
Technology should not replace people.
It should multiply human capability, time, and freedom.
Mission
Build reliable, professional, and elegant AI infrastructure so everyone can safely use the intelligence that fits them best, wherever they are, at a fair cost.
Vision
Intelligence, as natural to access as power and the internet
Become the intelligent foundation people rely on every day, freeing them from devices, locations, and any single model so more time returns to creating, deciding, and living.
Reliable · Keep every promise
A promise must hold. Prepare for critical moments and never pass avoidable uncertainty on to the user.
Professional · Steward complexity
Prepare, govern, and use resources wisely. Keep compatibility, routing, and cost complexity behind the infrastructure.
Long-term · Build with deep roots
Dig deep, plant patiently, and refine the foundation without chasing short-term noise.
Frequently Asked Questions
What is RABAH.AI?
RABAH.AI is AI infrastructure connecting developers, applications, and different models. One OpenAI-compatible API provides access to leading vendors with unified routing, monitoring, and usage management, without locking you into a single model or ecosystem.
Which models does RABAH.AI support?
RABAH.AI connects the latest models from OpenAI, Anthropic, Google, xAI, and more (such as GPT, Claude, Gemini, and Grok), with new models onboarded continuously.
How do I integrate with RABAH.AI?
Get an API key and point your base_url to https://api.rabah.ai. RABAH.AI provides compatible entry points for Chat, Responses, Claude Messages, and Gemini GenerateContent, so existing applications and agent tools can migrate with minimal changes.
How does RABAH.AI protect my data privacy?
By default RABAH.AI does not store request or response bodies — only token counts are recorded for billing. Chat also ships with built-in end-to-end encryption (E2EE), so sensitive conversations stay encrypted end to end.
How does smart routing work?
RABAH.AI selects a better-fit path based on model capability, cost, latency, and availability, helping you balance reliability, quality, and cost without hard-coding model selection into your application.