Syntic Code model configuration

Syntic Code runs on Syntic’s own models, and your plan decides which ones you can choose. Type /model in a session to see the list for your account and switch; the list comes from the Syntic platform, so it stays current without an update.

The models

Model	Context window	Plans
Syntic Code (default)	256K tokens	All paid plans
Syntic Code Fast	256K tokens	All paid plans
Amara Deep	256K tokens	All paid plans
Amara S2	1M tokens	All paid plans

Every paid plan currently includes all four models. If a model is ever limited to higher plans, /model shows it with the plan it needs.

Ways to set the model

How	Lasts
/model (or Alt+P), pick a row, press Enter	Saved as your default
/model, pick a row, press s	This session only
syntic --model "Syntic Code Fast"	This run
"model" in a settings file, or the SYNTIC_MODEL variable	Every session that reads it
/fast (or Alt+O)	Turns fast mode, which uses Syntic Code Fast, on or off

/effort sets how much the model reasons before answering (low to max). Lower effort is quicker and uses fewer tokens.

If a model is missing or refused

What happens	Why	What to do
The model is not in /model	Your organization’s policy hides it.	Ask your administrator.
The request is refused and names a plan	Your plan does not include that model.	Pick another model, or upgrade.
”The service stayed busy … through several retries”	The model was overloaded.	Wait a minute, or switch model. You can list backup models with "fallbackModel".

Learn more

• Models in the Syntic Code docs
• Plans and models