Unverified50% confidenceFactExact time
An LLM inference project called Nori claims to have achieved a processing throughput of over 1 million tokens per second (1M tok/s).
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
9/24/2026
First Seen
Valid until: 12/23/2026
Sources
Related Entities
Related Claims
UnverifiedUnder this configuration, the 35B model achieves a measured single-user peak speed of approximately 80 tokens/second, with concurrent throughput exceeding 500 tokens/second.66% similarUnverifiedThe inference speed of mainstream cloud-based large models currently typically ranges from tens to one or two hundred tokens per second.66% similarVerifiedA single inference at the million-token level requires tens to hundreds of gigabytes of high-end GPU memory (A100/H100 class).65% similarUnverifiedllama.cpp通过纯CPU推理和4-bit GGUF量化格式，使7B参数模型在无GPU的普通服务器上达到每秒20-40 tokens的生成速度64% similarUnverifiedThe Ornith 9B model weights occupy approximately 6.2 GB and can support a context of approximately 186,000 tokens within a 12 GB memory budget64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/947007API
curl https://kongchang.com/api/v1/knowledge/claims/947007MCP
get_claim(id=947007)