Provider pricing

Zyphra AI API pricing

Zyphra builds Zamba2 hybrid Mamba+Transformer models with linear-time inference. Pricing is tracked alongside hundreds of other providers in the TokenCost directory so teams can compare cost, capabilities, and tradeoffs in one place.

Founded2023
HeadquartersPalo Alto, CA
Tracked modelsDirectory

Best for

Edge and on-device deployments needing fast long-context inference.

Strengths

  • Hybrid Mamba+Transformer arch
  • Linear-time long-context inference
  • Open weights under Apache 2.0
  • Listed in the TokenCost AI providers directory