Verified65% confidenceFactExact time
Mistral AI's Mixtral 8x7B adopts a MoE architecture with a total parameter count of approximately 47B, but only approximately 13B parameters are actually activated during each inference.
3
Sources
65%
Confidence
Long-term
Relevance
7/13/2026
First Seen
Sources
AI Costs Out of Control: Real-World Strategies for Enterprises to Rein In Soaring Bills
hackernewshackernews7/11/2026
Related Claims
UnverifiedIn the MoE architecture, the Mixtral 8x7B model has a total of 46.7B parameters, but only approximately 12.9B parameters (2 experts) are activated per token, making inference costs close to those of a 13B dense model while achieving performance comparable to the 70B level.77% similarUnverifiedMistral AI的7B模型在发布时性能超越了两倍参数量的Llama 2 13B75% similarVerifiedMixtral 8x7B模型拥有8个专家，每个Token只激活2个，总参数量达到47B，但单次推理的计算量仅相当于一个13B参数的稠密模型75% similarPartially VerifiedMistral Large 4 adopts a sparse Mixture of Experts (MoE) architecture, with 49 billion (49B) parameters actually activated during each inference.71% similarUnverifiedThe actual total parameter count of Mixtral-8x7B is 46.7B, while the naming multiplication of 8x7=56B would mislead hardware planning by about 20%.71% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/501325API
curl https://kongchang.com/api/v1/knowledge/claims/501325MCP
get_claim(id=501325)