Skip to content

From Chat Completions to the Responses API

The OpenAI Responses API is the new standard for language models. Here is how to migrate existing integrations – with short examples for A2Intelligence.

A2Genesis team
A2Genesis GmbH · Bochum
1 min read
Generated dot-pattern cover image for the article

The Responses API replaces Chat Completions as the standard interface. It simplifies requests, cleanly separates reasoning from the answer and makes tool calls more robust.

The difference in one example

Instead of a list of messages you pass an input:

bash
curl https://gateway.a2genesis.de/v1/responses \  -H "Authorization: Bearer $A2_API_KEY" \  -H "Content-Type: application/json" \  -d '{"model": "a2intelligence", "input": "Summarise this contract in three points."}'

Do I have to migrate right away?

No. A2Intelligence still supports /v1/chat/completions as a legacy endpoint. For new projects we recommend the Responses API.

All articles
A2Genesis team

We build A2Intelligence, operated in Germany, and support businesses with software development, AI integration and automation.

More articles

6 min read

AI image styles (LoRA) explained

What does an AI image style do, what is behind LoRA and what should you know about licences? A clear introduction with examples from the starter set of styles.

6 min read

GDPR and AI: a checklist for businesses

How can your business use AI tools in line with the GDPR? DPA, server location, training, retention periods and documentation as a checklist you can tick off.

Trying it out takes
a few minutes.

Create an API key on the platform and test A2Intelligence directly with the Responses API.