Unverified50% confidenceBenchmarkExact time
The TEF 0.8B version model is only 811MB in size; in actual testing, running an entire set of classification tasks with TEF 0.8B took only 0.23 seconds, while the same tasks with GPT took about 2.5 seconds.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/12/2026
First Seen
Valid until: 10/26/2026
Sources
Related Entities
Related Claims
UnverifiedA 2B model requires approximately 4–10 TFLOPs of computation per inference, taking 10–50 milliseconds on an A100 GPU, while a 70B model requires 140–350 TFLOPs and can take up to several seconds.78% similarUnverifiedOpenAI's GPT-OSS 20B achieved a measured speed of over 21 tokens/second, nearly twice that of the dense 12B model.71% similarUnverifiedTaking a 70B BF16 model as an example, generating one token requires reading approximately 140 GB of weights, and when fully utilizing the H100's 3.35 TB/s memory bandwidth, the per-token latency is at least 42 milliseconds, with a theoretical TPS upper limit of approximately 24 tokens/second.71% similarUnverifiedGPT-OSS-20B BF16 loading was reduced from 545 seconds to 70 seconds (7.8× speedup), and Qwen3.5-397B achieved a 1.93× to 2.3× improvement.68% similarUnverifiedOn the RTX 4070 SUPER, OpenDLSS NR takes 7.8 milliseconds per frame at 1080P, 12.6 milliseconds at 2K, and 29.3 milliseconds at 4K.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1010873API
curl https://kongchang.com/api/v1/knowledge/claims/1010873MCP
get_claim(id=1010873)