Skip to content
AskVault Docs
Try Free →

Model selector

Written by Joseph Russo, Principal Architect, Networks and Tech, AskVault · Reviewed by Allison Chan
Last updated: · 3 min read

Default model per plan

Auto-assigned:

  • Free. Lightweight model (Gemini Flash). Fast, capable, no credit-cost tiers to manage.
  • Starter, Growth, Business, Enterprise. No plan-locked default: pick any model from the catalog (see below). Premium models cost more message credits per answer than lightweight ones.

Choosing a model (any paid plan)

For Starter workspaces and above:

  1. Settings > Chat Model.
  2. Pick from available list. Each model shows its message-credit cost per answer before you confirm.
  3. Save.

The new model answers the next reply.

Comparing models

Use Compare Models in Chat Playground:

  • Run the same question against 2 to 4 models.
  • Compare the answers side by side, with each model's credit cost.
  • Pick based on your priority.

Cost vs quality trade-offs

Roughly, in ascending order of message-credit cost per answer:

  • Lightweight (Free tier default). Fastest, lowest cost. Good for simple FAQs.
  • Mid-tier. Better at multi-step reasoning. Costs more credits per answer than lightweight.
  • High-capability. Best for complex queries, long context. Highest credit cost per answer.

Any paid plan can pick from the full range: the trade-off is credits per answer, not which plan you're on. For high-volume bots: prefer mid-tier unless quality requires higher.

Quick replies use a fast model

Short, simple messages, like a greeting, a "thanks" or a one-line question, are answered by a fast, lightweight model so the reply arrives quickly. This applies only when the message has no images and no skills are enabled. Longer or multi-part questions always use the model you selected.

Once fast-lane billing is switched on (it is rolling out and not yet enabled), a quick reply costs the fast model's price (1 credit), not your selected model's. Until then, quick replies use your selected model's credit price.

Limits

  • Available models. Fixed lightweight model on Free; full catalog on Starter and above (cost gated by message credits, not plan tier).
  • Granularity. One chat model per workspace.

Common pitfalls

Switched to a lower model, quality dropped. Switch back in the model selector, or tighten the system prompt for the lighter model.

Latency jumped after model switch. Higher-capability models often slower. Trade-off.

Cost spike. Higher-tier model means more tokens per query. Audit usage post-switch.

FAQ

Can I see which model is currently active?

Yes, the model selector in Settings shows it.

Will switching models lose conversation context?

No. Past messages preserve; the next turn uses the new model.

Does switching affect cost immediately?

Yes. Next query bills at new model rate.

Was this page helpful?