Frequently Asked Questions
Everything you need to know about CheapAI, our pricing, usage, and guarantees.
General & Ordering
After you place an order, the checkout shows your unique order number, exact payment amount, destination wallet, and live payment status. Once the blockchain payment is confirmed, your API Key + Proxy URL are sent to your email within 24 hours — usually much faster.
Bitcoin, Ethereum, USDT, USDC, BNB, XRP, Solana, TRON, and Litecoin. Always send on the exact network shown — wrong network means permanent loss.
No. We only accept cryptocurrency payments. This ensures your privacy and avoids recurring subscription traps or unexpected charges.
Yes. If you run out of tokens or want to extend your time-based plan, simply contact us on Telegram (@cheapai1sell) with your order ID, and we can top up your account or issue a new key immediately.
Yes. CheapAI provides access through its own infrastructure and delivery flow. You receive real API credentials and a compatible base URL that work with standard tools and frameworks.
Pricing & Plans
Choose your model, set your budget ($10–$2,000), and you get output tokens at 50–65% off official prices. The more you buy, the higher the discount.
The Claude Pack gives you time-based access to all Anthropic models (Opus 4.6 Thinking, Sonnet 4.6, Haiku 4.5) with a usage credit limit. Pick daily, weekly, or monthly plans from the dropdown.
Our goal is to keep prices as low as possible. We combine discounted sourcing arrangements, lean operations, and community-scale purchasing to keep rates down. If our base cost changes materially, we aim to give active users advance notice where practical.
Every plan is guaranteed for its full duration. If your API key stops working — we replace it immediately or issue a full refund. Contact @cheapai1sell on Telegram.
If your API key stops working during your plan period, contact @cheapai1sell on Telegram. We'll either replace your key immediately or issue a full refund — no questions asked. Your plan is 100% guaranteed for its full duration.
Technical & Integration
Yes. Any tool with a custom base URL works — Cursor, OpenWebUI, LangChain, LlamaIndex, Continue.dev, and most coding frameworks.
Response speed is designed to stay close to the underlying provider experience, but real latency still depends on model load, request size, and network conditions. The proxy layer is meant to stay lightweight, not to change the model behavior.
We use the exact same APIs you'd connect to, directly from the provider. There is no middle-man AI or filtering. The GPT-5 or Sonnet 4.6 you use is identical to the official model—just cheaper.
No. The proxy handles all routing for you. You can access the API from any country worldwide without needing a VPN on your end.
Absolutely! System prompts, temperature, Top_P, max_tokens, and all other standard OpenAI-compatible JSON parameters are passed directly to the model without alteration.
Yes. Streaming is fully supported and enabled by default across all our endpoints. You will see characters appear in real-time just like ChatGPT or Claude's official UI.
Yes. Models that support images and documents (like GPT-5 and Claude 4.6) fully support image inputs via standard base64 encoding or image URLs through our proxy.
The service is built for real developer workloads, including scripting, automation, and batch usage. Fair-use and infrastructure protections may still apply when needed to keep the platform stable for everyone.
Yes. You can share your key across Cursor on multiple machines, your custom web apps, and your backend scripts simultaneously without any conflict.
Privacy & Support
CheapAI is designed to keep retained data as limited as possible. Short-lived technical processing or temporary optimization layers may be used where required for delivery, stability, or abuse prevention, but the public site does not rely on long-term browser-side storage of your API keys or checkout data.
Use the service according to your own security and compliance requirements. We aim to minimize retained data and keep the public checkout flow lean, but teams handling highly sensitive information should always apply their own internal review and vendor-risk standards.
CheapAI acts as the access layer for the supported services you buy from us. That means your personal account on an official provider is not required for the public purchase flow shown on this site. As with any third-party service, you should still use the product in a lawful and policy-compliant way.
Currently, we do not have a formal affiliate system set up, as our margins are incredibly tight. However, if you are an influencer or can bring high volume, message us on Telegram for a custom arrangement.
We operate asynchronously with team members in multiple timezones. The fastest way to get support is always via Telegram (@cheapai1sell), where we monitor messages closely.
Setup Guides & Pricing Comparisons
Step-by-step documentation for the most common CheapAI setups:
- Cheap Claude API — Claude Opus 4.6 + Sonnet 4.6 pricing comparison
- Cheap GPT API — GPT-5.4 + Codex pricing comparison
- Cheap Gemini API — Gemini 3 Pro + Flash pricing comparison
- OpenAI-Compatible API guide — drop-in replacement for any OpenAI SDK
- Cursor AI setup guide — use Claude & GPT inside Cursor
- Claude Code setup guide — full CLI walkthrough
- All models directory — Claude, GPT-5, Gemini, DeepSeek in one place
Page maintained by the CheapAI team. Last updated: March 2026. About CheapAI →