Not our models just yet.

Powerful at its core ...

A perfect blend of Powerful and Free, to offer the optimal choice in any situation, with an easy pick from 3 great Models.
... when needed.
Eos 0
On-Device
Fast and Free, even offline. Ada with Eos makes local AI easy, it makes all the hard decisions for you, and only lets you decide what really matters.
Parameter:
2B - 31B
Context:
82k - 256k
Speed:
•••••
Accuracy:
•••••
Best for:
Everyday questions, and quick help.
Powered by:
Gemma 4
Helios 0
Cloud
Sharper reasoning, longer context, and powerful enough for most of your Tasks. Perfect as your everyday Powerhouse.
Parameter:
119B
Context:
256k
Speed:
•••••
Accuracy:
•••••
Best for:
Research, writing, documents, code.
Powered by:
Mistral Small 4
Hyperion 0
Cloud
Our most Powerful model, for even the heaviest Tasks, that deserve nothing but the best.
Parameter:
1.05T
Context:
1M
Speed:
•••••
Accuracy:
•••••
Best for:
Complex analysis, long projects, hard decisions.
Powered by:
Mistral Large 4

Pay for what you use.

Model Input Cached input Output Searches
Eos 0On-Device
 Free Free Free €TBA
Helios 0Cloud
 €TBA €TBA €TBA €TBA
Hyperion 0Cloud
 €TBA €TBA €TBA €TBA
Prices per 1M tokens.
No subscriptions, for now. We think subscriptions are a good fit for AI and hope to offer one in the future, but at launch it's pay as you go only. With prices this low, you shouldn't end up paying much more.