Pricing & Plans

Two Flexible Ways to Pay for AI Compute

Use your own API keys (BYO-Tokens) and pay only for gateway relay, or top up your Aitok balance to pay directly for model consumption.

Choose either method or mix both seamlessly.

BYO-Tokens

Key Benefit

Use Your Own Keys (BYO-Tokens)

Zero Compute Advance

Draws from your existing provider balance — no double-paying

Connect your API keys from OpenAI, Anthropic, Google Gemini, Meta Llama, Mistral, and more. Aitok handles intelligent failover routing, access controls, and usage observability.

Your own keys, your own provider billing
Compatible with all major AI providers & open weights
Millisecond-relay proxy with real-time analytics

You pay for

0.00000002 t / req (First 3,000 reqs / mo free)

Managed Compute

Key Benefit

Aitok Unified Compute (Managed)

No Provider Accounts Needed

One balance, access all major models & agents instantly

Top up your Aitok balance directly and invoke any supported model or agent on-demand — zero individual account or payment card hassles.

No individual provider signups required
Unified 1:1 token-pegged balance
Instant turnkey access to official & community models

You pay for

Actual Model Consumption

Unified routing & governance via Aitok Core Gateway

All-in-One AI Compute

Managed, routed, and secured by Aitok Gateway

Gateway API Docs →
BYO-Tokens Mode

Bring Your Own API Keys (BYO-Tokens)

Already hold provider balances across OpenAI, Anthropic, Google Gemini, Mistral, etc.? Connect them to Aitok and pay a nominal gateway fee only for excess requests, with partner models completely free.

Your Existing Provider Keys

Own Balance
OpenAI Key
Connected
Anthropic Key
Connected
Google Gemini Key
Connected
Meta Llama Key
Connected
Mistral AI Key
Connected
Cohere Key
Connected

Aitok Unified Gateway

High-concurrency auth & metered relay

Intelligent Multi-Model Routing
Real-Time Throughput Monitoring
Team Key Access Isolation

AI Agents & Clients

Plug & Play
Claude Code
CLI Agent
Cursor
AI Code Editor
Cline
VS Code Extension
Continue.dev
IDE Copilot

Gateway Relay Billing Tiers

Includes 3,000 free relay requests each month; pay only a nominal flat fee for excess volume

→
03,000 reqs / mo∞
Monthly Free AllowanceMetered Excess Requests

Included Every Month

3,000 Free Requests / Month

3,000reqs / mo100% Free

A generous monthly allowance of 3,000 free requests, covering everyday development and debugging.

Resets on the 1st of every month

Beyond Free Allowance

Pay-As-You-Go Excess Requests

0.00000002 tper single request

Pay only for requests exceeding the free monthly quota. Partner models incur zero relay fees.

Transparent flat-rate with zero hidden minimums

Bring your own keys. The first 3,000 requests each month are completely free, with additional requests metered at transparent micro-rates.

Keep Your Own Provider Discounts
3,000 Free Requests Monthly
Multi-Protocol & Provider Support
Centralized Cross-Client Key Management
Managed Compute Mode

Use Aitok Unified Compute Balance

Directly access all supported AI models through Aitok. Top up once and access models from multiple providers without managing separate accounts or balances.

Step 1

One Balance

Aitok WalletTop Up Once

Balance

1,000 t

Ready to Use
Top Up →

1 t corresponds 1:1 with base currency units, with transparent real-time auditability.

Step 2

Access All Supported AI Models

GPT-4oOpenAI
Claude 3.5 SonnetAnthropic
Gemini 1.5 ProGoogle
Llama 3.3 70BMeta
Mistral LargeMistral
o1-miniOpenAI

No need to fund multiple accounts. One unified balance accesses 30+ flagship models.

Step 3

Use Anywhere

Claude Code
Terminal Agent
Cursor
AI IDE
Cline
Coding Assistant
Continue.dev
IDE Copilot
Chatbox
Desktop App
Workbuddy
Team Workspace

Seamlessly connects with CLI tools, IDEs, desktop workspaces, and AI agent frameworks.

One balance. All supported AI models.

Top up once through Aitok to access AI models from multiple providers without managing separate accounts.

1
One balance
2
All supported models
3
All AI experiences

Pay-As-You-Go

No subscriptions. No minimums. Pure consumption-based.

Per-Model Pricing

Different AI models consume tokens at different rates.

Exact Metering

Billed precisely by actual AI usage.

Direct AI Access
Pay Only for What You Use
All Flagship Models Supported
Real-time Usage Tracking
Capability

Hybrid Supply· Use Both. Manage in One Place.

Already have your own AI tokens? Need access to more models through Aitok? Use both side by side in the same workspace.

Use What You HaveSupplement What You NeedUnified Management

Your Tokens

OpenAI
Claude
+

Aitok Tokens

Gemini
Meta Llama

Start Powering Your AI Stack Today

Get started in 3 minutes. Connect your own provider keys or top up to access instant AI compute.

Monthly Free Gateway Requests
Automated Multi-Model Failover
Multi-Protocol Compatible

Own API Keys

+

Aitok Compute Pool

Aitok Core Gateway
Apps · Agents · CLI