OpenAI in AGNT
GPT-class models for general reasoning, tool use and code. Every AGNT workflow is provider-portable, so the same agents run here or anywhere else you point them.
Why teams pick it
Models
- GPT-5.2
- GPT-5 mini
- o-series reasoning
Strengths
- strongest ecosystem and tool-calling conventions
- reliable structured output
- broad task competence
In AGNT
Agents, workflows and tools use OpenAI like any other provider — same nodes, same receipts, swap models per step if you want.
When OpenAI is the right default
It is the safest starting point for agent work because the tool-calling conventions everything else imitates originated here. Structured output is dependable, the reasoning tier handles genuinely hard multi-step problems, and almost every prompt technique you will read about was developed against these models first. If you have no strong reason to choose otherwise, start here and move once you have a measured reason.
What to watch
Cost accumulates quickly on agent loops that re-send a large prefix every turn, so prompt caching matters more here than the pricing page suggests. Reasoning models also trade latency for quality in a way that is excellent for a nightly analysis job and poor for an interactive step, which is an argument for routing per step rather than committing the whole workflow to one tier.
Using OpenAI in a workflow
In practice you rarely commit a whole workflow to one model. A common arrangement puts GPT-5.2 on the steps where strongest ecosystem and tool-calling conventions genuinely decides the outcome, and something cheaper and faster on the routine classification around it. Because AGNT normalises tool calling, that split is configuration per step rather than two separate builds.
Connect OpenAI in two minutes
- Download AGNT Community Core — free, local-first, no account needed to run.
- Open Settings → Providers, choose OpenAI, and paste your API key. Connect by pasting an API key into AGNT’s vault — stored encrypted on your machine, never uploaded.
- Give an agent a job — or install one from the marketplace — and watch the first receipt come back.
OpenAI in AGNT — common questions
Which OpenAI model should an agent use?
A mid-tier model for routine steps and a reasoning model reserved for the genuinely hard call. Routing per step rather than per workflow is usually the largest cost lever available.
Do I need an OpenAI account to use AGNT?
No. AGNT is free and runs locally; you can point it at a local model through Ollama and never create a provider account at all.
Can I switch away later without rebuilding?
Yes. AGNT normalises tool calling across providers, so the model behind a step is configuration rather than architecture.