Providers
Providers
Providers are the AI services that Quotio connects to. These are the companies or platforms that host the AI models you use.
What is a provider?
In Quotio, a provider represents a connection to an external AI service or platform that hosts language models. Each provider connection includes:
Provider Components
- Service Endpoint: The API server URL (e.g.,
api.openai.com,api.anthropic.com) - Authentication: API keys, OAuth tokens, or service account credentials
- Model Catalog: Available AI models and their capabilities
- Quota Limits: Rate limits, spending caps, and usage tracking
- Billing Integration: Cost calculation and usage reporting
Supported Providers
Quotio integrates with CLIProxyAPI/CLIProxyAPIPlus to support a wide range of AI providers through OAuth authentication:
Core CLI Providers (via CLIProxyAPI):
- Gemini CLI - Google's Gemini models via CLI authentication
- Antigravity - Antigravity's AI models
- ChatGPT Codex (OpenAI) - OpenAI's Codex models via OAuth
- Claude Code - Anthropic's Claude models via CLI authentication
- Qwen Code - Alibaba's Qwen models via CLI authentication
- iFlow - iFlow's AI models via CLI authentication
Additional Providers (via CLIProxyAPIPlus):
- GitHub Copilot - GitHub's Copilot service via OAuth
- Kiro (AWS CodeWhisperer) - Amazon's CodeWhisperer via OAuth
Enterprise/Direct Providers:
- Vertex AI (Google Cloud) - Service account JSON authentication
- Ollama - Local model inference (no authentication required)
Provider Characteristics
Each provider has unique characteristics that affect how Quotio manages it:
- Pricing Model: Per-token, per-request, or subscription-based
- Rate Limits: Requests per minute (RPM) and tokens per minute (TPM)
- Context Window: Maximum input + output token length
- Model Families: Different model sizes and capabilities
- API Compatibility: OpenAI-compatible vs. proprietary APIs
Provider vs. Agent Distinction
Providers are the AI services themselves, while Agents are the CLI tools that use those providers. For example:
- Provider: "OpenAI API"
- Agent: "Claude Code CLI tool that calls OpenAI API"
This separation allows one provider connection to serve multiple agents.
Why providers matter
Different providers have different strengths:
- Claude: Excellent for code generation and analysis
- GPT-4: Versatile for general tasks
- Gemini: Good for multimodal (text + images)
- Local models: Zero cost but limited capabilities
Quotio lets you use multiple providers simultaneously for maximum flexibility.
Provider types
Quotio supports providers through different authentication methods:
Authentication Methods
All providers in Quotio authenticate through the CLIProxyAPI backend using different methods:
OAuth Authentication (Browser-based login):
- Gemini CLI: Opens browser for Google OAuth
- Claude Code: Opens browser for Anthropic OAuth
- Qwen Code: Opens browser for Alibaba OAuth
- iFlow: Opens browser for iFlow OAuth
- GitHub Copilot: Device code flow via browser
- Kiro (AWS CodeWhisperer): AWS OAuth authentication
- ChatGPT Codex: OpenAI OAuth authentication
- Antigravity: Google's Antigravity IDE OAuth
Service Account Authentication (JSON file upload):
- Vertex AI: Google Cloud service account JSON
No Authentication (Local or self-hosted):
- Ollama: Local model inference (no authentication)
Connecting providers
In the Providers tab:
- Click Add Provider
- Select your provider
- Follow the authentication flow:
- OAuth: Browser popup for login
- API Key: Paste your key
- Service Account: Import JSON file
Once connected, Quotio starts tracking quota immediately.
Provider status
Each connected provider shows:
- Status: Ready, Error, or Rate Limited
- Quota: Current usage percentage
- Models: Available AI models
- Last Used: When it was last accessed
Managing multiple providers
With multiple providers connected:
- Load balancing: Distribute requests across providers
- Failover: Auto-switch when one provider is exhausted
- Cost optimization: Route to cheapest available model
Next steps
Learn how Quotio tracks usage across all your providers.