← Back to CreditFX

SambaNova Cloud

Inference platform (open models) · LLM APIs

Representative rate
$0.6/M out tok

Custom silicon (not GPUs) is the pitch — genuinely fast inference on open models, priced competitively with Groq and Cerebras.

All pricing tiers

API (Llama 3.3 70B)custom · $0.6/M out tok

Only one tier verified so far — more being added over time.

Want to calculate exact usage instead of just the representative rate? Open the full calculator.