llmapi FAQ: Models, Tools, Plans and Payment

What is llmapi.cheap?

llmapi.cheap gives you one API key for top AI models, built for coding agents. The key works with OpenAI-compatible tools (Chat Completions and Responses at https://llmapi.cheap/v1), Anthropic-compatible tools (Messages at https://llmapi.cheap) and the Gemini API. You pay a flat monthly price instead of per token.

Which coding tools work with it?

A single setup command configures Claude Code (CLI, VS Code, JetBrains), Codex (CLI, IDE, app), OpenCode, the Claude desktop app, GitHub Copilot in VS Code, Kilo Code, Cline, Roo Code, Continue, Hermes Agent, OpenClaw, Crush, Factory Droid, Qwen Code, Gemini CLI, Aider, Goose, Junie CLI, Raycast AI, Chatbox and Cherry Studio. Cursor, Zed, Visual Studio, JetBrains AI Assistant, Trae and Warp are set up by hand in their settings screen (in Cursor, chat works; Cursor's own agent features need Cursor's servers). Windsurf, Cursor CLI and Background Agents, Augment Code, Kiro and Amazon Q can't be connected because they only talk to their makers' servers. See the setup guides.

How do I set it up?

Sign up and an API key (starting with sk-llm-) is created automatically. Then run one command. On macOS or Linux: curl -fsSL https://llmapi.cheap/install.sh | sh. On Windows PowerShell: irm https://llmapi.cheap/install.ps1 | iex. You don't need Node.js installed. The dashboard shows the same command with your key already embedded. The script backs up every file it changes and only adds llmapi's own settings.

Which models are available?

Gemini 3.8 Flash with low, medium and high reasoning (gemini-3.8-flash-low, gemini-3.8-flash, gemini-3.8-flash-high), Gemini 3.1 Pro (gemini-3.1-pro-high, gemini-3.1-pro-low), Gemini 3 Flash (gemini-3-flash), Claude Sonnet 4.6 (claude-sonnet-4-6), Claude Opus 4.6 (claude-opus-4-6) and GPT-OSS 120B (gpt-oss-120b). Gemini 3.8 Flash (high) is the default everywhere. In Claude Code, /model lists all of them.

Is there a free trial?

Yes. Every new account gets 2M credits on all Gemini models, valid for one hour from your first request, not from sign-up. No card or wallet is needed. Claude and GPT-OSS models need a paid plan.

How much does it cost?

Four monthly plans: Lite $2, Plus $5 (the most popular), Pro $8 and Max $12. Lite includes 12M Gemini credits per 5 hours and 36M per week, plus 0.6M Claude & GPT-OSS credits per 5 hours and 1.5M per week. Plus includes 21M Gemini per 5 hours and 63M per week, plus 1M Claude & GPT-OSS per 5 hours and 2.5M per week. Pro and Max include more of both. Every plan has unlimited parallel requests. See pricing.

How do the usage limits work?

Your first request opens a 5-hour window, and you can use up to your plan's allowance inside it. When the window closes, the next request starts a fresh one. A weekly cap limits the total over seven days. Gemini and Claude & GPT-OSS are metered separately. Larger models use credits faster, and cached input counts at 10% of normal input, so long sessions with stable context go further. When you hit a limit, the error tells you when it resets. Read more in running a coding agent for $2–$12 a month.

How do I pay?

In USDT (Tether). Pick a plan in the dashboard and send the exact amount shown on the invoice to the address shown, on the network shown. The main network is BNB Chain (BEP20), where fees are a few cents. Your plan starts automatically, usually about a minute after the transfer confirms. Buying the same plan again extends it; buying a different plan replaces the current one.

Can I get a refund?

For refunds or any billing issue, please contact support. We recommend using the free trial first to make sure llmapi works with your tools.

Do you see or train on my code?

Your requests are forwarded to the model provider to generate a response. llmapi does not train on them. For your usage page we record token counts, the model and latency, not the content of your prompts and responses.

What does the dashboard show?

Your current 5-hour and weekly usage, usage by model, your API keys and your plan. It also shows the real dollar value of what you used at official API prices, while you only pay the plan price.

Do I need to re-run the setup after upgrading?

No. The setup adds every model to your tools' model lists from the start, so Claude and GPT-OSS start working as soon as your plan is active. Re-running it is harmless if you want to.

Create a free account, run the one command, and start coding. The free hour is enough to try llmapi on your own project.