RDU-accelerated Llama 3 with very high tok/s
SambaNova Cloud — RDU-accelerated Llama 3 with very high tok/s.
Free tier: Free developer tier with rate-limited access (10 RPM typical, varies by model). Frontier-scale 405B available on free tier — rare.
API is OpenAI-compatible — point your SDK at https://api.sambanova.ai/v1.
Meta-Llama-3.3-70B-InstructMeta-Llama-3.1-405B-InstructMeta-Llama-3.1-8B-InstructDeepSeek-R1Higher rate-limit dev tiers and dedicated capacity (contact sales). Llama 3.1 405B priced at $5/1M tokens (input) on paid tier.
On-prem SambaNova systems, sovereign cloud, financial-services tier.
https://api.sambanova.ai/v1
OpenAI-compatible — works with the OpenAI SDK by overriding base_url.
curl https://api.sambanova.ai/v1/chat/completions \
-H "Authorization: Bearer $SAMBANOVA_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "Meta-Llama-3.3-70B-Instruct",
"messages": [{"role": "user", "content": "Hello in 5 words"}]
}'
Trying Llama 3.1 405B for free. Apps where speed of large open models matters more than ecosystem.
Closed models or vision; only Llama family + DeepSeek today.
Free developer tier with rate-limited access (10 RPM typical, varies by model). Frontier-scale 405B available on free tier — rare. No credit card is required to sign up.
Most commonly used: Meta-Llama-3.3-70B-Instruct, Meta-Llama-3.1-405B-Instruct, Meta-Llama-3.1-8B-Instruct, DeepSeek-R1. The full current list is on SambaNova Cloud's docs page.
Yes — point the OpenAI SDK's base URL at `https://api.sambanova.ai/v1` and pass your SambaNova Cloud API key.
Higher rate-limit dev tiers and dedicated capacity (contact sales). Llama 3.1 405B priced at $5/1M tokens (input) on paid tier. If your traffic is bursty or seasonal, the free tier may be enough; if you need a guaranteed SLA, upgrade.