Direct OpenAI-compatible API access exclusively for GPT-5.6 Sol, Terra & Luna models. 100% authentic provider compute, flexible daily/weekly unlimited plans or API credits.
Official configuration files and test commands for Windows, Linux, and WSL.
{
"OPENAI_API_KEY": "ta-live-key-your-key-here"
}Choose between unlimited token compute passes or pay-as-you-go API credit packages for GPT-5.6 models.
Lightweight agent runs & personal dev
2 Active Concurrent Connections
Zero per-request token caps
GPT-5.6 Sol, Terra & Luna
For high-frequency multi-file coding agents
4 Active Concurrent Connections
Zero per-request token caps
GPT-5.6 Sol, Terra & Luna
For continuous swarm runs & team workloads
6 Active Concurrent Connections
Zero per-request token caps
GPT-5.6 Sol, Terra & Luna
Upgrade anytime mid-subscription by paying only the price difference.
Prices are final. Plans are refundable PRO RATA in the event of cancelled services.
Legit official provider models—zero proxy downgrades or fake model swaps.
We strip away token counting and proxy degradation, offering clean, direct API pipes for coding agents.
No extra rerouting, silent model swapping, or quantized proxies. The model you request is the exact model responding.
Manage scoped credentials, restrict access by IP range, rotation policies, and export environment variables instantly.
Send long contexts, full file context trees, and large diffs without getting hit by arbitrary token cap errors.
Drop-in replacement for OpenAI SDKs, LangChain, LlamaIndex, Cursor, VS Code, and Codex CLI by altering only the base URL.
Compare official metered OpenAI token rates against TorpiloAPI's unthrottled daily passes.
Fully compliant with the OpenAI REST specification. Use standard headers and endpoints.
Pass your API key as a Bearer Token in the HTTP Authorization header:
Authorization: Bearer ta-live-your-api-key-here
Base Endpoint: https://torpiloapi.com/v1
POST https://torpiloapi.com/v1/chat/completions
{
"model": "gpt-5.6-sol",
"messages": [
{"role": "system", "content": "You are an expert software engineer."},
{"role": "user", "content": "Implement a binary search tree in Rust"}
],
"temperature": 0.2
}Query GET https://torpiloapi.com/v1/models to inspect available high-performance models:
gpt-5.6-sol — High-throughput flagship reasoning & coding model.terra — Ultra-fast model optimized for IDE inline autocomplete & diff generation.luna — High-precision multimodal & deep context analysis model.| Code | Status Name | Description / Resolution |
|---|---|---|
401 |
Unauthorized | Invalid API key or inactive plan / credit balance. Contact support on Discord. |
429 |
Rate Limit Exceeded | Concurreny limit reached (e.g. > 2 or > 4 active sessions). Retry with exponential backoff. |
500 |
Provider Fail | Transient upstream network retry. Automatically routed to secondary buffer. |
Works natively with Cursor, VS Code, Codex CLI, Claude CLI, Hermes, and OpenAI SDKs.
Set Custom OpenAI Base URL to https://torpiloapi.com/v1 in Cursor Settings.
Configure config.json with apiBase: "https://torpiloapi.com/v1".
Configure ~/.codex/config.toml and ~/.codex/auth.json with TorpiloAPI base URL.
Pass base_url="https://torpiloapi.com/v1" into official OpenAI SDK client instances.
You receive direct OpenAI-compatible API access credentials (API Key and Base URL https://torpiloapi.com/v1) for GPT-5.6 Sol, Terra & Luna. Choose between Unlimited Token Passes (M, L, XL Daily or Weekly passes with 2, 4, or 6 simultaneous API calls) or Pay-As-You-Go API Credit Packs ($5 for 100 Credits up to $37 for 1,000 Credits).
Yes. 100% authentic official provider models—zero proxy downgrades, silent model swapping, or quantized proxies. The model you request is the exact model responding.
Prices are final. Plans are refundable PRO RATA in the event of cancelled services.
Yes! Account upgrades are possible even within an active subscription—you only pay the price difference to get upgraded to the next tier immediately.
Credit Card (+5%) and PayPal (+5%) payments incur a standard processor fee. Crypto payments on the BSC network carry 0% extra processing fee.