Provider pricing

SambaNova AI API pricing

SambaNova Cloud runs the largest open-weight models — Llama 4 Maverick, DeepSeek V4, Qwen3 — on its Reconfigurable Dataflow Unit (RDU) hardware, delivering high throughput at full BF16 precision (no quantization). The free developer tier includes generous rate limits and an OpenAI-compatible API.

Founded2017
HeadquartersPalo Alto, CA
Tracked models1

Best for

Quality-sensitive production workloads on flagship open-weight models.

Strengths

  • Full BF16 precision — no quality loss from quantization
  • Strong throughput on the largest open-weight models
  • Free developer tier with production-grade rate limits
  • Composition of Experts: route across multiple specialist models

Tracked SambaNova models

ModelInput $/MOutput $/MContextTierBest for
Llama 3.3 70B (SambaNova)$0.600$1.20128,000Mid-tierRDU-accelerated Llama 3.3 70B for fast enterprise inference