- Context
- 131K
- Price / 1M
- $0.13 out
$0.03 in
$0.02 cached - Precision
- 4 bits
Model library
Models
Every model below runs on a processor configured for that model — not on general-purpose silicon it has to share. Same OpenAI-compatible API, same model weights, same accuracy.
8 models
- Context
- 131K
- Price / 1M
- $0.30 out
$0.08 in
$0.07 cached - Precision
- 4 bits
- Context
- 262K
- Price / 1M
- $1.94 out
$0.08 in
$0.06 cached - Precision
- 4 bits
- Context
- 262K
- Price / 1M
- $0.91 out
$0.07 in
$0.06 cached - Precision
- 4 bits
- Context
- 1M
- Price / 1M
- $0.25 out
$0.01 in
$0.01 cached - Precision
- 4 bits
- Context
- 262K
- Price / 1M
- Coming soon
- Precision
- 4 bits
- Context
- 262K
- Price / 1M
- $3.98 out
$0.22 in
$0.18 cached - Precision
- 4 bits
- Context
- 262K
- Price / 1M
- $0.47 out
$0.04 in
$0.03 cached - Precision
- 4 bits
Don’t see the model you run?
The processor is rebuilt around the model, so onboarding a new architecture is a compile step — not a new chip. Tell us what you’re serving and we’ll bring it up.