Groq in AGNT
LPU-served open models at extreme speed. Every AGNT workflow is provider-portable, so the same agents run here or anywhere else you point them.
Why teams pick it
Models
- Llama on LPU
- open-weight menu
Strengths
- lowest latency serving
- high throughput for agent loops
- simple flat pricing
In AGNT
Agents, workflows and tools use Groq like any other provider — same nodes, same receipts, swap models per step if you want.
Latency as the product
Groq serves open-weight models on custom hardware at speeds that change what interactive agent work feels like. The gain compounds in loops: a workflow making twelve sequential calls turns a noticeable wait into something that feels immediate, which matters more for user-facing steps than a marginal quality difference would.
What to watch
You are choosing from a menu of open models rather than a proprietary frontier model, so capability is bounded by what is available. The right use is high-volume routine steps and interactive loops, with harder judgment calls routed elsewhere in the same workflow.
Using Groq in a workflow
In practice you rarely commit a whole workflow to one model. A common arrangement puts Llama on LPU on the steps where lowest latency serving 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 Groq in two minutes
- Download AGNT Community Core — free, local-first, no account needed to run.
- Open Settings → Providers, choose Groq, 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.
Groq in AGNT — common questions
When is Groq the right choice?
Interactive steps and high-frequency loops where latency dominates the experience.
Which models can I run on it?
A menu of open-weight models. Check current availability rather than assuming a specific checkpoint.
Can one workflow use Groq and a frontier model?
Yes, and it is the recommended pattern: speed where speed matters, depth where depth matters.