Overview
ZeroLimit supports simultaneous connections to multiple AI service providers, enabling you to manage all your AI quotas, track usage, and switch between accounts seamlessly from one unified dashboard.Supported Providers
ZeroLimit integrates with six major AI providers:Antigravity
Multi-model AI router supporting Claude, GPT, and Gemini models with advanced quota management
OpenAI Codex
ChatGPT Plus and Free tier access with model-specific quota tracking
Gemini CLI
Google AI Studio and Vertex AI with project-based quota buckets
Kiro (CodeWhisperer)
AWS AI coding assistant with tiered usage limits
GitHub Copilot
GitHub’s AI pair programmer with subscription-based quotas
Claude (Anthropic)
Direct Anthropic API access with usage monitoring
Provider Details
Antigravity
Type: Multi-model AI routerAuthentication: OAuth
Project ID: Not required
Features
- Access to Claude Sonnet 4.5, Opus 4.5, thinking models
- Gemini 3 Pro, Gemini 2.5 Flash, Flash Lite
- Gemini image models
- Grouped quota display by model family
- Average usage calculation across model groups
Model Groups
- Claude/GPT: High-tier reasoning models
- Gemini 3 Pro: Premium Google models
- Gemini 2.5 Flash: Fast generation models
- Gemini Image: Visual AI models
OpenAI Codex
Type: ChatGPT APIAuthentication: OAuth
Project ID: Not required
Features
- ChatGPT Plus and Free tier support
- GPT-4, GPT-4 Turbo, o1 models
- GPT-3.5 Turbo variants
- Per-model quota limits
- Plan detection (Plus/Free/Suspended)
Quota Display
- Shows remaining requests or message credits
- Reset times for each model
- Plan-specific limits highlighted
Codex tracks both Plus subscription limits and free tier caps. ZeroLimit automatically detects your plan.
Gemini CLI
Type: Google AI Studio / Vertex AIAuthentication: OAuth with optional Project ID
Project ID: Optional (auto-detected if omitted)
Features
- Project-based quota buckets
- Multiple Google Cloud projects supported
- Per-project quota tracking
- Bucket-level reset times
- Model-specific rate limits
Use Cases
- Development projects with separate quotas
- Multi-tenant applications
- Testing vs. production environments
If you have multiple Google Cloud projects, add Gemini CLI connections for each project to track quotas separately.
Kiro (CodeWhisperer)
Type: AWS AI coding assistantAuthentication: OAuth (email or SSO)
Project ID: Not required
Features
- Individual and Professional tier support
- Monthly usage quotas
- IDE integration tracking
- Plan status monitoring
Authentication Methods
- Email-based OAuth
- AWS SSO
- Multiple accounts per user
GitHub Copilot
Type: AI pair programmerAuthentication: GitHub Device Code Flow
Project ID: Not required
Features
- Subscription-based quota tracking
- Individual, Business, Enterprise plans
- Usage statistics across IDEs
- Real-time availability status
Authentication Flow
- Click GitHub Copilot provider
- Copy the displayed device code
- Open GitHub authorization page
- Enter code and approve
- Connection completes automatically
Claude (Anthropic)
Type: Direct Anthropic APIAuthentication: OAuth
Project ID: Not required
Features
- Claude 3 Opus, Sonnet, Haiku models
- Thinking mode variants
- Token-based quota limits
- Per-model usage tracking
- Organization-level quotas for teams
Model Tracking
- Separate quotas for each Claude model version
- Context window limits monitored
- Rate limits displayed with reset times
Multiple Accounts per Provider
ZeroLimit allows multiple accounts for the same provider:Use Cases for Multiple Accounts
- Personal + Work: Separate personal and professional subscriptions
- Free + Paid: Keep free tier for testing, paid for production
- Backup: Redundancy when one account hits quota limits
- Project-Based: Different accounts for different clients or projects
Provider Comparison
| Provider | OAuth | Plus Only | Project ID | Multi-Account |
|---|---|---|---|---|
| Antigravity | ✅ | ❌ | ❌ | ✅ |
| OpenAI Codex | ✅ | ❌ | ❌ | ✅ |
| Gemini CLI | ✅ | ❌ | Optional | ✅ |
| Kiro | ✅ | ✅ | ❌ | ✅ |
| GitHub Copilot | ✅ Device Code | ✅ | ❌ | ✅ |
| Claude | ✅ | ❌ | ❌ | ✅ |
Cross-Provider Features
Unified Dashboard
All providers share the same monitoring interface:- Consistent quota card design
- Standardized color coding (green >20%, yellow ≤20%)
- Reset time display
- Privacy mode support
Provider Filtering
Quickly filter by provider:- Click provider tabs at top of Quota page
- View only accounts for selected provider
- Provider count badge shows number of accounts
- Switch instantly between providers
Bulk Operations
Perform actions across all providers:- Download All: Export all auth files as ZIP
- Copy All Tokens: Selectively copy refresh tokens by provider
- Delete All: Remove all accounts at once
Icon System
Each provider has a distinct icon for visual recognition:- Antigravity: Purple gradient logo
- Codex: OpenAI circle logo
- Gemini CLI: Google Gemini sparkle
- Kiro: CodeWhisperer icon
- Copilot: GitHub Copilot logo (auto-inverts in dark mode)
- Claude: Anthropic Claude logo
Icons automatically adapt to light/dark themes for optimal visibility.
Quota Aggregation
Model Grouping
ZeroLimit intelligently groups models by family:Average Calculation
For each group:- Individual model percentages are summed
- Divided by number of models in group
- Rounded to nearest integer
- Displayed with group icon and color coding
Best Practices
Diversify Providers
Diversify Providers
Connect multiple providers to avoid single points of failure. If one service is down or quota is exhausted, switch to another.
Name Conventions
Name Conventions
Use consistent email naming for personal vs. work accounts (e.g.,
[email protected] vs. [email protected]).Monitor Plus-Only Providers
Monitor Plus-Only Providers
If using Standard CLI Proxy, note which providers require Plus and plan your upgrade accordingly.
Project Organization (Gemini)
Project Organization (Gemini)
For Gemini CLI, use project IDs to separate dev, staging, and production quotas.
Adding New Provider Support
ZeroLimit’s architecture supports easy addition of new providers. The CLI Proxy API handles:- OAuth endpoints and callback routing
- Token refresh logic
- Quota API integration
- Provider-specific metadata
Provider-Specific Quota APIs
Each provider uses a different quota API:- Antigravity: Custom management API with model buckets
- Codex: OpenAI backend API with account credits
- Gemini CLI: Google Cloud quota buckets by project
- Kiro: AWS usage metrics endpoint
- Copilot: GitHub internal usage API
- Claude: Anthropic account dashboard API
Troubleshooting
Provider Won’t Connect
- Verify CLI Proxy is running
- Check if provider service is operational
- Clear browser cache and retry OAuth
- Ensure correct version (Standard vs. Plus)
- Check firewall allows localhost:8317
Quota Not Updating
- Click refresh button on individual provider card
- Verify internet connection
- Check if account is suspended on provider website
- Try deleting and re-adding the account
- Review CLI Proxy logs for API errors
Plus-Only Provider Shows Lock
Switch to CLI Proxy Plus version:- Go to Settings → CLI Proxy
- Click Switch to Plus
- Wait for download and restart
- Return to Providers page and connect
Next Steps
Quota Monitoring
Learn how to track and analyze your multi-provider quotas
Proxy Management
Manage the CLI Proxy server and provider connections