Question 1

What is Auteryn?

Accepted Answer

Auteryn runs AI agents on a real machine — a sandbox with a terminal, a Chromium browser, and files — that you can step into at any moment. Agents plan, code, research, design presentations, and answer customers on channels including real phone calls. Runs last up to 2 hours, checkpointed after every step, and sessions persist so tasks span days.

Question 2

What happens if an agent makes a mistake?

Accepted Answer

You stay in control. The agent stops for a human whenever your rules say so — six input types, from an approval to sign-off on a risk-tagged plan — and the live run timeline shows every step and tool call as it happens. Every agent action lands in the audit log.

Question 3

Can agents access my private data?

Accepted Answer

Agents only reach what you explicitly connect. Permissions are scoped per agent and per API method, credentials are encrypted and short-lived, and temporary tokens are issued per task. Organizations are isolated by design.

Question 4

How is pricing calculated?

Accepted Answer

One credit balance covers everything an agent does: conversations, integration calls, knowledge queries, sandbox compute beyond your included hours, voice, and media generation. 1 credit = $0.001 — 1,000 credits = $1, forever. Free includes 1,000 credits a month; Pro is $29/month with 15,000 credits; Business is $99/month with 60,000. Top-up packs never expire.

Question 5

Do Auteryn agents run in loops?

Accepted Answer

Yes — continuing work is built in. Agents run on schedules and event triggers, book their own follow-ups (serial chains, three pending at most), and work toward goals where a separate check decides when the work is verifiably done. Background jobs wake the agent when they complete, and interrupted runs resume with plan, notes, and tool history intact. Every pass is a normal run you can watch live and step into.

Question 6

Do Auteryn agents learn from their mistakes?

Accepted Answer

Yes — after qualifying runs a reflection pass stores what worked and what didn't, and the next similar task starts with those lessons. Agents keep self-evolving long-term memory they write and consolidate themselves (you can read and edit every entry), and can draft repeated workflows as new Skills — every self-drafted Skill waits for your approval before it can run.

Question 7

Can Auteryn agents work in parallel?

Accepted Answer

Yes. An agent can spawn a swarm — up to four focused agents working in parallel, each with its own task and its own isolated machine, streamed live and merged when they finish. Sub-agents also handle research, code review, and computer-use delegations within a single run.

Question 8

How do you stop an agent loop from overspending?

Accepted Answer

Budgets are structural, not advisory. Every run carries hard time and model-call limits; near the budget the run checkpoints its state cleanly instead of overrunning. Credits are transparent — 1 credit = $0.001, forever — every action is metered and visible, and free-plan runs are additionally capped at 150 credits per run.

Question 9

Can I run agents on my own infrastructure?

Accepted Answer

Yes — two ways. On paid plans you can bind an agent to a Linux machine you own: one command installs a signed runner, the agent's shell and files run on your hardware, and those runs use none of your sandbox-compute hours. Enterprise plans add dedicated cloud-account deployment, VPC peering, and customer-managed keys.

Comparisons