Satelink combines model routing, context, tools, execution controls, and usage visibility. Start free or choose an execution tier tailored to your workflow.
Controlled discovery tier for testing and evaluation.
500 UU/wk (~50 tasks)
Refills up to 100 UU per 5h session
Normal AI assistance, research, and moderate agent workflows.
+ applicable taxes (18% GST in India)
7,500 UU/wk (~750 tasks)
Refills up to 1,500 UU per 5h session
Substantially higher execution allowance for frequent users and power workflows.
+ applicable taxes (18% GST in India)
37,500 UU/wk (~3,750 tasks)
Refills up to 7,500 UU per 5h session
Heavy research and continuous multi-step autonomous agent execution.
+ applicable taxes (18% GST in India)
150,000 UU/wk (~15,000 tasks)
Refills up to 30,000 UU per 5h session
Statutory disclosure: All subscription amounts are billed upfront. Annual plans reflect twelve months upfront at approximately 16.7% discount versus twelve single-month renewals. Prices are exclusive of applicable statutory taxes (18% GST applies to domestic Indian billing).
We reject arbitrary markup on provider tokens and opaque message caps. Execution is measured in verifiable Usage Units (UU) corresponding directly to underlying computational costs.
Input and output tokens are routed at raw upstream list prices. Fast routing uses lightweight models (~0.5 to 1 UU/task); complex synthesis uses Claude 3.5 Sonnet (~10 to 25 UU/task); multi-step deep reasoning is metered with explicit task ceilings.
External research queries, web fetches, document parsing, and Python code sandbox execution draw modest, deterministic units. Failed executions and retries are capped at 3 attempts with exponential backoff to prevent runaway bills.
Bring your own API keys from Anthropic, OpenAI, or Google. Upstream inference is billed directly to your provider account; Satelink draws zero token markup, billing only orchestration, structured memory, and tool execution.
Inspired by modern frontier AI platforms, Satelink uses a dual-window enforcement architecture to protect your weekly allowance while guaranteeing high-speed execution bursts.
Refills continuously over a rolling 5-hour window. This ensures a single heavy research task cannot accidentally deplete your entire weekly allowance in a burst.
Your full weekly allowance resets cleanly every Monday at 00:00 UTC. Clear progress bars in your console track exact consumption in real time.
If your session or weekly allowance is reached, Satelink notifies you with clear options: wait for the rolling refill, upgrade your tier, draw from purchased usage credits, or switch to BYOK routing. Nothing is silently billed.
Pro Plan Weekly Window (7,500 UU allowance)
Hard spend limits are checked pre-flight before executing multi-turn tool or agent calls.
Need heavy continuous execution without upgrading your plan? Add non-expiring usage credits that draw only when your included subscription allowance is depleted.
Custom amounts available from ₹500 to ₹250,000 in your Console. Payment processed via Transacty (UPI / cards) and PayRam (non-custodial crypto).
Every feature is verified through actual backend enforcement. Planned and private beta features are labeled honestly.
| Feature | Free | Pro | Max 5× | Max 20× | Enterprise |
|---|---|---|---|---|---|
| Core Execution Allowances | |||||
| Weekly Included Allowance | 500 UU (~50 tasks) | 7,500 UU (~750 tasks) | 37,500 UU (~3,750 tasks) | 150,000 UU (~15,000 tasks) | Custom pooled quota |
| Session Burst Window | 100 UU / 5 hours | 1,500 UU / 5 hours | 7,500 UU / 5 hours | 30,000 UU / 5 hours | Dedicated burst capacity |
| Reset Schedule | Every Monday 00:00 UTC | Every Monday 00:00 UTC | Every Monday 00:00 UTC | Every Monday 00:00 UTC | Custom cycle / contract |
| Credit Pack Auto-draw | |||||
| Uninterrupted Execution Rollover | Purchased credits only | Purchased credits only | Purchased credits only | Custom rollover agreement | |
| Model Intelligence & Routing | |||||
| Standard Fast Models (Flash, Haiku, Llama 3.3) | |||||
| Workhorse Models (Claude 3.5 Sonnet, GPT-4o) | |||||
| Deep Reasoning Passes / Month | 5 passes | 25 passes | 60 passes | Custom / Unrestricted | |
| Priority Traffic Routing at Peak Hours | Highest queue priority | Isolated private compute | |||
| Bring Your Own Key (BYOK) Pass-through | Single key | Up to 5 keys | Up to 20 keys | Unlimited organization pooling | |
| Tools, Concurrency & Workspaces | |||||
| Live Web Search & Document Extraction | |||||
| Secure Sandbox Code Execution | |||||
| Market Microstructure Research Feeds | Sample queries | Full metered access | Full metered access | Priority metered access | Dedicated websocket feeds |
| Maximum Concurrent Agent Tasks | 1 task | 3 tasks | 10 tasks | 25 tasks | Custom scale |
| Active API Keys | 1 key | 5 keys | 20 keys | 50 keys | Custom / Unlimited |
| Persistent Structured Memory | Session only | 5 projects | Unlimited projects | Unlimited + vector store | Dedicated tenant memory |
| Governance, Budgets & Controls | |||||
| Deterministic Hard Spend Caps | |||||
| Usage Notifications (70%, 85%, 95%, 100%) | In-app | Email & Webhooks | Email & Webhooks | Email & Webhooks | Custom incident triggers |
| Circuit Breaker on Repeated Tool Errors | |||||
| Detailed Consumption Breakdown by Model/Tool | Summary | Daily logs | Full telemetry | Real-time stream | Audit export |
| Single Sign-On (SSO) & SCIM Directory | |||||
| Commercial, Rails & Support | |||||
| Fiat Payment Rails (Cards / UPI via Transacty) | N/A | Invoicing / Net 30 | |||
| Crypto Rails (USDT on Polygon, USDC on Base via PayRam) | |||||
| Statutory Tax Invoicing (GST compliant) | N/A | ||||
| Support Tier | Community Docs | Standard Email | Priority Email (24h) | Dedicated Queue (4h) | Dedicated Engineer & Slack |
Everything you need to know about Satelink billing, taxes, allowances, and payment rails.