On the web
Every plan includes all three sizes. Auto is the default; switch in any conversation.
Models
Three sizes. One contract.
Fast, Balanced and Deep are stable ids you can depend on. The model behind a size can change; the size, its context window and its price are the contract. Auto picks one for each request.
Fast· MiraQuick answers and everyday work, with a 400ms latency target.64K context
- Input / 1M
- $0.20
- Output / 1M
- $0.40
- Input / 1M
- $0.65
- Output / 1M
- $1.30
- Input / 1M
- $0.76
- Output / 1M
- $1.52
Auto
Auto picks by the size of the request. You can always choose.
Auto is the default everywhere: on the web, in Cruxy Code and through the API. It reads how long and complex a request is and routes it to one size.
Short promptsFast· Mira
Long or tool-heavy requestsDeep· Kavi
Everything elseBalanced· Vaani
Compare
Side by side.
SIZE
Fast
Mira
Balanced
Vaani
Deep
Kavi
Context window
64K
128K
256K
Max output
2K
4K
8K
Modalities
Text
Text
Text
Input, per 1M tokens
$0.20
$0.65
$0.76
Output, per 1M tokens
$0.40
$1.30
$1.52
Rates are USD per 1M tokens, the rates API usage is metered at. API prices exclude GST. 18% GST is added for India-billed accounts (input credit available). On the web, every plan includes all three.
Everywhere
The same three sizes, wherever you work.
In Cruxy Code
Auto by default. Pin a size with /model for the session, in .cruxy/config.json for a project, or with CRUXY_MODEL.
Through the API
Pass auto, or kavi, vaani or mira by id. You pay the rate of the model that served the call.
Pick a size. Or let Auto.
Free on the web, no card required. Per token through the API.