What is SambaNova Turbo
SambaNova Turbo is represented by SambaNova Cloud, a high-speed inference service with production and preview models, API keys, a playground and usage-based developer billing.
Free, Developer and Enterprise
As of August 16, 2026, Free provides pay-as-you-go credits, production-model access and community support. Developer is pay as you go with standard rate limits and access to production and preview models. Enterprise uses subscription pricing for larger workloads, production limits and standard support, with add-ons for on-request models, bring-your-own-cloud and custom limits.
Current model pricing
Per million tokens, MiniMax-M2.7 costs US$0.60 input, US$0.06 cached input and US$2.40 output. DeepSeek-V3.1 and V3.2 each cost US$3 input and US$4.50 output. Gemma 4 31B costs US$0.38 and US$1.15, GPT OSS 120B costs US$0.22 and US$0.59, and Meta Llama 3.3 70B Instruct costs US$0.60 and US$1.20. These are all models on the current official table; absent or retired models are excluded.
Official sources: SambaNova Cloud plans, SambaNova model pricing.