News
ReleasesOVHcloud unveils Olala mini, its first in-house LLMChinaA 9B model beats a 1T model on FinQA, trained with Crusoe Managed RLReleasesThe internal metaphysics of an LLMReleasesMapping Language Model Internal Representations to Physical Laws and Scientific ReasoningResearchCan China keep rogue AI in check? Meet the start-up testing the top modelsFrontierMicrosoft aims to challenge OpenAI, Anthropic with cheaper, specialized AI modelsOpen modelsMicrosoft rolls out new first-party models and updates to agentic AI continuous improvementFrontierMicrosoft unveils AI models for coding, cybersecurity as it takes on Anthropic, OpenAIFrontierGoogle is making it easier to opt out of Gemini AI model training – here’s howFrontierOpenAI Unveils New Astra AI Model After Scrapping Prior ReleaseReleasesOVHcloud unveils Olala mini, its first in-house LLMChinaA 9B model beats a 1T model on FinQA, trained with Crusoe Managed RLReleasesThe internal metaphysics of an LLMReleasesMapping Language Model Internal Representations to Physical Laws and Scientific ReasoningResearchCan China keep rogue AI in check? Meet the start-up testing the top modelsFrontierMicrosoft aims to challenge OpenAI, Anthropic with cheaper, specialized AI modelsOpen modelsMicrosoft rolls out new first-party models and updates to agentic AI continuous improvementFrontierMicrosoft unveils AI models for coding, cybersecurity as it takes on Anthropic, OpenAIFrontierGoogle is making it easier to opt out of Gemini AI model training – here’s howFrontierOpenAI Unveils New Astra AI Model After Scrapping Prior Release
EveryLLM

Mixture of experts

Mixtral, DeepSeek-V3, Kimi K2: many expert networks, a few used for each token. 112 models on EveryLLM so far.

A mixture of experts (MoE) model replaces each dense feed-forward block with many smaller expert networks and a router that sends each token to only a few of them. The model can hold a very large number of parameters while using only a fraction for any one token, so it learns more without costing proportionally more to run. Mixtral, DeepSeek-V3, Qwen3, Llama 4, gpt-oss and Kimi K2 are all built this way.

Total and active parameters

An MoE has two sizes. DeepSeek-V3 (December 2024) holds 671 billion parameters but activates 37 billion for each token. Mixtral 8x7B (December 2023) has eight experts per layer, uses two for each token, and has 46.7 billion parameters in total with 12.9 billion active. Kimi K2 (July 2025) reaches one trillion in total with 32 billion active. All the experts must still sit in memory, so MoE models need large servers even when they are fast.

A long history

The idea dates to 1991, and Google researchers led by Noam Shazeer revived it for language in 2017 with a sparsely gated layer of thousands of experts. Google's Switch Transformer reached 1.6 trillion parameters in 2021. Open MoE models took off with Mixtral, then Databricks' DBRX (March 2024, 132 billion total, 36 billion active), and became the norm for large Chinese open models. Training them is harder, because routers must spread tokens evenly across experts.

Famous mixture of experts models

Further reading

Solar Open 100B 🇰🇷 South KoreaSolar Open 100B 102B parameter Mixture-of-Experts model by UpstageGeneral · Conditional licence · 2026 fnSolar Open 2 🇰🇷 South KoreaSolar Open 2 250B parameter open-weight model for agentic tasksGeneral · Open weights · 2026 fnKanana-2 🇰🇷 South KoreaKanana-2 Open-source multilingual model for agentic AI and reasoning.General · Open weights Qwen3.8-2.4T-A95B 🇨🇳 ChinaQwen3.8-2.4T-A95B Open-weight reasoning model with 2.4 trillion parametersReasoning · Open weights · 2026 imgvidfnKimi K2.6 🇨🇳 ChinaKimi K2.6 Open-source multimodal agentic model by Moonshot AIGeneral · Open weights fnKimi K2 🇨🇳 ChinaKimi K2 Open mixture-of-experts model for agentic tasks and codingGeneral · Open weights · 2025 imgfnKimi K3.1 Pro 🇨🇳 ChinaKimi K3.1 Pro Open-weight multimodal agentic model with 1M token contextMultimodal · Open weights fnGLM-4.5 🇨🇳 ChinaGLM-4.5 Open-source mixture of experts model for agents and reasoningGeneral · Open weights · 2025 Kimi Linear 🇨🇳 ChinaKimi Linear Hybrid linear attention model with 48 billion parametersGeneral · Open weights · 2025 imgfnLing-1T 🇨🇳 ChinaLing-1T Trillion-parameter efficient reasoning model by InclusionAIGeneral · Open weights imgLongCat-2.5 🇨🇳 ChinaLongCat-2.5 Multimodal MoE model for long-horizon agentic tasksMultimodal · Open weights fnLongCat-2.0 🇨🇳 ChinaLongCat-2.0 A 1.6 trillion parameter MoE language model from Meituan.General · Open weights MiniMax-Text-01 🇨🇳 ChinaMiniMax-Text-01 A large language model with a four million token context window.General · Open weights · 2025 fnMiniMax-M1 🇨🇳 ChinaMiniMax-M1 Open-weight reasoning model with hybrid attention and long contextReasoning · Open weights · 2025 fnLongCat-Flash 🇨🇳 ChinaLongCat-Flash Efficient 560B parameter MoE model for agentic tasksGeneral · Open weights · 2025 wavfnSarvam-30B 🇮🇳 IndiaSarvam-30B Open-weight MoE model for Indian languages and codingGeneral · Open weights PanGu-Sigma 🇨🇳 ChinaPanGu-Sigma Trillion-parameter sparse language model for Chinese NLP tasksGeneral · Research only · 2023 fnGLM-5 🇨🇳 ChinaGLM-5 Open-weight model for engineering and agentic tasksGeneral · Open weights · 2026 fnERNIE-4.5-21B-A3B-Thinking 🇨🇳 ChinaERNIE-4.5-21B-A3B-Thinking Lightweight reasoning model with 21 billion parametersReasoning · Open weights · 2025 imgERNIE 4.5 🇨🇳 ChinaERNIE 4.5 Baidu's multimodal mixture-of-experts language model series.General · Open weights · 2025 fnHunyuan-A13B 🇨🇳 ChinaHunyuan-A13B Open-source MoE language model with 80 billion total parametersGeneral · Open weights · 2025 Hunyuan-TurboS 🇨🇳 ChinaHunyuan-TurboS Hybrid Mamba-Transformer MoE model by TencentGeneral · API only · 2025 Seed1.5-Thinking 🇨🇳 ChinaSeed1.5-Thinking A reasoning model from ByteDance Seed using reinforcement learningReasoning · API only · 2025 imgStep3 🇨🇳 ChinaStep3 Multimodal reasoning model with 321 billion parametersMultimodal · Open weights · 2025 imgvidIntern-S1 🇨🇳 ChinaIntern-S1 Open-source multimodal reasoning model for scienceReasoning · Open weights · 2025 MiMo-V2.5-Pro 🇨🇳 ChinaMiMo-V2.5-Pro Open-source MoE language model with 1M token contextGeneral · Open weights imgwavvidMiMo-V2.6-Pro 🇨🇳 ChinaMiMo-V2.6-Pro Xiaomi’s 1T parameter multimodal model with 1M token contextMultimodal · Open weights fnStep-3.5-Flash 🇨🇳 ChinaStep-3.5-Flash Open-source MoE model for reasoning and agentsGeneral · Open weights · 2026 MAI-1-preview 🇺🇸 United StatesMAI-1-preview Microsoft AI's first in-house foundation model, a mixture-of-experts system for everyday queriesGeneral · API only · 2025 imgvidNemotron 3.5 Super VL 🇺🇸 United StatesNemotron 3.5 Super VL NVIDIA's open-weight vision-language model with 120B parameters and 1M-token contextMultimodal · Conditional licence · 2026 imgvidDiffusionGemma 🇺🇸 United StatesDiffusionGemma A multimodal Gemma 4 model that writes text by denoising blocks of tokens.Multimodal · Open weights Qwen2 🇨🇳 ChinaQwen2 Open-weight language model series from AlibabaGeneral · Open weights · 2024 imgfnLeanstral 1.5 🇫🇷 FranceLeanstral 1.5 An open code agent model built for the Lean 4 proof assistantCode · Open weights fnNorth Mini Code 🇺🇸 United StatesNorth Mini Code A 30B sparse mixture-of-experts model for code and terminal agent workCode · Conditional licence fnGranite 4.0 🇺🇸 United StatesGranite 4.0 IBM's hybrid Mamba/transformer language models for enterprise workGeneral · Open weights · 2025 imgfnMistral Large 3 🇫🇷 FranceMistral Large 3 A 675B-parameter open-weight mixture-of-experts multimodal model from Mistral AIGeneral · Open weights imgMM1 WorldMM1 A family of multimodal language models up to 30B parameters, dense and MoE.Multimodal · Research only · 2024 imgApple Foundation Models 🇺🇸 United StatesApple Foundation Models Apple's 2025 on-device and server language models for Apple IntelligenceGeneral · API only · 2025 Mixtral 8x22B 🇫🇷 FranceMixtral 8x22B Sparse mixture-of-experts open model with 39B active parameters out of 141BGeneral · Open weights fnGpt-oss-120b 🇺🇸 United StatesGpt-oss-120b OpenAI's open-weight reasoning model that fits on one 80GB GPUReasoning · Open weights · 2025 fnGpt-oss-20b 🇺🇸 United StatesGpt-oss-20b OpenAI's smaller open-weight reasoning model for local and low-latency useReasoning · Open weights · 2025 imgwavvidGemini 1.5 Pro 🇺🇸 United StatesGemini 1.5 Pro Google DeepMind's mid-size multimodal model with up to a 1 million token context windowGeneral · API only · 2024 imgwavvidGemma 4 🇺🇸 United StatesGemma 4 Google DeepMind's open multimodal models in five sizes, from phones to serversGeneral · Open weights Switch Transformer WorldSwitch Transformer A sparse mixture-of-experts language model scaling to a trillion parameters.Research · Research only · 2021 GLaM 🇺🇸 United StatesGLaM A trillion-parameter mixture-of-experts language model from Google ResearchResearch · Research only · 2021 NLLB-200 🇺🇸 United StatesNLLB-200 A research translation model covering 200 languages, built for low-resource ones.General · Conditional licence · 2022 imgPhi-3 🇺🇸 United StatesPhi-3 Small Microsoft language models built to run well on phones and laptops.Small · Open weights · 2024 DeepSeek-V3 🇨🇳 ChinaDeepSeek-V3 Open-weight mixture of experts language model with 671 billion parameters.General · Open weights · 2024

The weekly digest

Every Monday: the model of the day, new pages on EveryLLM, and the conferences and anniversaries of the week ahead.