Skip to main content

Mind pricing isn't discoverable before a chat begins

Published Minds can carry a creator markup on inference, flagged on listings with a "Premium" badge. The markup percentage, the underlying model, and its rate are all absent from anywhere a prospective user can see before sending a message. Cost is revealed neither on the Details page not in the Composer (they way it would be with a pay-per-use model in Agentic Chat). Instead there is merely “Premium Mind. Messages can use credits.” which could be anything from Sonnet to Fable. Creator Markup only compounds this problem.

This means users only find out how much the premium minds are costing them when they go to their API/usage pages, possibly after hours of use.

Steps to reproduce

  1. Open any listed Mind carrying the "Premium" badge.

  2. Read the details page.

  3. Observe that the page says the Mind is pay-per-use, but does not name the model, its rate, or the markup.

Observed

  • Listing card: badge reads "Premium" (paired with Anon or Private). No figure.

  • Details page: pay-per-use indicated. Model unnamed. No rate. No markup.

  • Chat: “Premium Mind. Messages can use credits”.

  • Edit view, Pricing panel: "Venice rate" is hardcoded to 100 Credits and does not change with the selected model, so even the creator can't see a live rate or their own markup.

Expected Enough to decide whether to send before sending. Minimum: the markup percentage. Ideally also the model and its rate.

Why it matters

  • The direction is inverted relative to the rest of the product. Raw models are priced pre-hoc — check the rate, then send. Minds are billed post-hoc — send, then find out.

  • The badge spans roughly an order of magnitude of cost, cheap models to frontier ones, so it can't be used to screen anything.

  • Users have to sample each Mind to learn its price, and pay to do so.

  • Creators are already working around it in copy — e.g. "Generate images using quota-based models — no credits required" — which shows the UI gap is real.

  • Transparency is important. While creators may desire their prompt choices to remain private, pricing always deserves transparency.

Suggested fixes, cheapest first

  1. Put the markup % in the card badge and on the details page (e.g. "Premium · +10%").

  2. Name the underlying model or model pricing on the details page.

  3. Show an estimated per-turn cost before sending in the composer

  4. Make the Pricing settings panel quote the selected model (for, say, 1000 tokens) instead of a fixed 100-credit placeholder.

  5. Show a running session total.

Log in to comment and vote

No comments yet

Be the first to share your thoughts.