Emvryx
Models Plans How usage works About Open Emvryx
HOW USAGE WORKS

You always know what is left.

Emvryx shows your usage in plain messages, right under the message box. There is a short-term amount for right now and a longer one for the week.

Right now

An E that fills like a tank shows how many messages you can send in one sitting. It refills by itself, steadily, so there is no waiting for a reset. From empty it is full again in about 8 hours.

This week

A weekly total, shown as a percentage and about how many messages are left. It resets every Monday at 4 AM in your own time zone. A strip of seven days shows which days were busy.

What a message costs

A message is one question and a normal-length answer. Bigger models use more of your usage. On a Pro week (about 1,000,000 tokens), that comes to roughly:

Ember
4,000
messages a week
Flare
2,000
messages a week
Equinox
1,000
messages a week
Solstice
400
messages a week
Apogee
200
messages a week

Thinking longer costs more

The effort setting next to the model chooser changes both the answer and the cost.

QuickFastest. Uses less: about 40% fewer.
BalancedNormal speed and normal usage.
DeepSlowest and most careful. Uses about twice as much.

Make it last

  • Long chats use more, because every message re-reads the chat so far. Start a new chat for a new topic.
  • Use Quick for simple questions.
  • Use a model on your own device: it is free and never counted.

For the curious

Behind the scenes, everything is counted in tokens (small pieces of words). The text you send counts once, earlier chat the server already holds counts at a tenth, and what Emvryx writes counts once, then it is multiplied by the model’s weight (Ember 0.5, Flare 1, Equinox 2, Solstice 5, Apogee 10). Tap “See the numbers” in the Usage screen to see them.

© Emvryx.AI 2026. Private by design.
ModelsPlansHow usage worksAbout