o10Last updated 2026-07-26

Squad

Multi-agent: planner → workers → judge. For hard multi-step work. Use model o10/squad.

Latency typically 1–4 minutes — not chat-speed. Beta / live.

Spread observed
638×
Routing modes
shadow → enforce
Framework
KYI
Dashboards observe.
o10 enforces.

Cost dashboards tell you what you spent. o10 sits in the request path and changes what you spend — shadow first, then enforce.

SummaryKey takeaways

What you need to know

Short, self-contained answers with cited stats — read the sections below for full context.

What is o10/squad?

Multi-agent: planner → workers → judge. For hard multi-step work. Use model o10/squad.

Beta / live

How fast is Squad?

Squad is typically 1–4 minutes — not chat-speed. It is built for hard multi-step work, not instant chat.

1–4 minutes typical

What is the default roster?

Default roster (informational; orgs can override in-app): planner/synth gpt-5.5, workers kimi-k2.5, judge/validator claude-sonnet-4.6.

01Deep dive

How Squad runs

One prompt in; roster receipt out.

Default roster (informational; orgs can override in-app): planner/synth gpt-5.5, workers kimi-k2.5, judge/validator claude-sonnet-4.6.

Open https://app.o10.io/squad after signup. Use model o10/squad with your free route key (placeholder $O10_ROUTE_KEY).

  • Planner breaks the work down
  • Workers run in parallel
  • Validator / judge checks quality
  • Synthesizer returns the final answer
How-toOperational steps

Get started

  1. 01

    Start free

    Sign up at https://app.o10.io/signup — free route key, no card.

  2. 02

    Open Squad

    Go to https://app.o10.io/squad in the product console.

  3. 03

    Set model

    Use o10/squad. Expect 1–4 minutes per run.

→Related topics

Continue across hubs, glossary, guides, and comparisons.

All modesFrontier TokensSmart routing aliasesStart free (no card)Open Squad in app
See what you're overpaying →
FAQFrequently asked questions

Common questions

What is Squad?

Multi-agent: planner → workers → judge. For hard multi-step work. Use model o10/squad. Default roster (informational; orgs can override in-app): planner/synth gpt-5.5, workers kimi-k2.5, judge/validator claude-sonnet-4.6.

How long does Squad take?

Squad latency is typically 1–4 minutes — not chat-speed. Plan for multi-minute runs, not chat-speed replies.

How do I run Squad?

Start free at https://app.o10.io/signup, open https://app.o10.io/squad, and set model to o10/squad. Status: Beta / live.

o10Set the envelope. o10 holds it.

See what you're overpaying.

Paste a week of traffic. Get the number that books the audit.

See what you're overpaying →
verified savings methodology · State of Inference Spend 2026