Search Results
2 results for “mixture-of-experts”
Models
Mixtral
Mixtral is a family of open-weight sparse mixture-of-experts large language models developed by Mistral AI, comprising Mixtral 8x7B and Mixtral 8x22B, released under the Apache 2.0 licence.
5 min readUpdated May 2026
Models
Qwen
Qwen is a family of large language models developed by Alibaba Cloud, ranging from small open-weight dense models to trillion-parameter mixture-of-experts systems, with strong multilingual and reasoning capabilities.
4 min readUpdated May 2026