Your tools, one gateway
Your own applications, Codex CLI and OpenCode use Logicc as a model gateway. One key, all available models.
For developers and automation
API
One API key for the leading models, via the Responses API with the official OpenAI SDK. Data protection and security come along with every call.
One endpoint. The leading models through one interface.
Compatible. The official OpenAI SDK works out of the box.
In every plan. Soon you can create API keys yourself in the product.
No training. Your data remains your data.
More than 3,000 companies work with Logicc
Overview
One endpoint for all models, without data risk.
one key
Claude
GPT
GeminiModel via parameter
GET /v1/models shows your models. Switching means changing one parameter.
The path of every request
Stored in Germany, processed in Europe, deleted at the model provider immediately.
Integrated in minutes
Official OpenAI SDK, Base URL api.logicc.com, done. Streaming and tool calling via the Responses API.
For developers The complete reference
The complete reference
in one place.
One interface, one key
OpenAI-compatible · Responses API
https://api.logicc.com/v1 - Auth
Authorization: Bearer $LOGICC_API_KEY- SDK
- Official OpenAI SDK with your own Base URL
- For
- Query models and generate responses
Access: included in every plan. Soon you can create API keys yourself directly in the product, without a request.
Endpoints
- GET
/v1/modelsAll models available for your key. Authoritative for valid model IDs.
- POST
/v1/responsesThe Responses API: text, documents, streaming and tool calling, depending on the model.
# pip install openai import os from openai import OpenAI client = OpenAI( base_url="https://api.logicc.com/v1", api_key=os.environ["LOGICC_API_KEY"], ) resp = client.responses.create( model="gpt-5.6-terra", input="Summarize the contract.", ) print(resp.output_text)
// npm install openai import OpenAI from "openai"; const client = new OpenAI({ baseURL: "https://api.logicc.com/v1", apiKey: process.env.LOGICC_API_KEY, }); const res = await client.responses.create({ model: "claude-sonnet-5", input: "Hello", });
curl https://api.logicc.com/v1/responses \ -H "Authorization: Bearer $LOGICC_API_KEY" \ -H "content-type: application/json" \ -d '{"model": "gemini-3.5-flash", "input": "Hello"}'
Pricing
Billing is per million tokens, depending on the model. Secure+ has its own prices. You can see the current prices per model directly in the platform; they are updated continuously.
Tools and applications
Wherever the OpenAI SDK or the Responses API is supported with your own Base URL.
- Your own applicationsBackend, scripts, internal tools · OpenAI SDK
Python, JavaScript - Codex CLITerminal · OpenAI Responses
Set Base URL - OpenCodeTerminal · OpenAI Responses
Set Base URL
Limits and go-live
Usage limit and technical rate limit are separate controls. A 429 can also occur below the budget. Recommended: limited retries with exponential backoff and jitter.
- Key in a secret manager, never in code
- Check Base URL api.logicc.com
- Confirm model IDs via /v1/models
- Timeouts and limited retries with backoff
- Validate tool inputs as valid JSON
- Logging without secrets
- Define fallback behavior
Agenda
How it works in three steps.
Key Create an API key
Create an API key
Soon directly in the product, in every plan and without a request.
Integration Check and connect models
Check and connect models
GET /v1/models shows your available models. Then: Responses API with the official OpenAI SDK.
Operation Automate securely
Automate securely
Streaming and tool calling included. Your data stays protected.
FAQ
Frequently asked questions.
Yes: storage in Germany, encrypted traffic, Zero Data Retention at the model providers, no training on your content.
With the official OpenAI SDK via the Responses API. Tools with your own Base URL, such as Codex CLI or OpenCode, can be connected directly. There is currently no Anthropic-compatible interface.
Billing is per million tokens per model. Secure+ has its own prices. You can see the current prices directly in the platform.
Every plan includes API access. Soon you can create API keys yourself directly in the product, without a request.
Usage limit and technical rate limit are separate. A 429 can also occur below the budget, so we recommend limited retries with exponential backoff.
The API documentation will get its own website. We will link it here as soon as it is online.
Get started securely with AI.
Try it free for 7 days. No credit card required. Or book a demo