When to use Copilot Auto instead of picking a model yourself?
Picking the right model per task is busywork that also inflates your bill. OpenCode's Copilot Auto option lets Copilot's router choose the model for each session at the Auto discount, and falls back to your account's default chat model if routing fails.
What it is
When your Copilot account has an Auto pool, you can pick Auto in OpenCode's model picker instead of a specific model. Copilot's router decides per session and OpenCode handles failure with a fallback.
Applied case
Imagine kicking off ten quick sessions across a day and hand-picking a model each time. A developer with a Copilot Auto pool instead selects Auto once in the model picker; Copilot's router assigns a model per session at the Auto discount. When routing fails on one session, OpenCode silently falls back to the account's default chat model, so work continues instead of erroring out. The router, not you, absorbs the decision.
- Pick Auto once in the model picker; Copilot routes per session and discounts it
When to use it
Use when your Copilot account has an Auto pool and you want cheaper routing without per-session model picks. Skip when you need a guaranteed specific model for a task, since Auto does not let you pin one, or when your account has no Auto pool.
Read more
Recommended for you
- VerificationVercel
When to use decision models instead of language models?
Teams burn budget and latency on LLMs for quick judgment calls like classification and scoring, when decision models do it at about 1/70th the cost per request.
- VerificationSuperpowersClaude Code
How do you stop agents from inventing tests that prove nothing?
An agent writes code, then writes a test for it. The test passes right away and proves nothing. Fix: no production code without a failing test first.
- VerificationSuperpowersClaude Code
Why is your agent's first fix attempt usually wrong?
An agent hits a bug, guesses a fix, reports done. The bug returns because the fix treated a symptom. Fix: no fixes without root cause.
- VerificationVercel
How do you catch vulnerabilities on direct pushes to production?
Direct and force pushes to production branches skip PR-based security checks entirely. Security Review scans every push and reports findings with suggested fixes.
Enjoyed this article?
Subscribe for new articles. No spam. Unsubscribe anytime.