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

Dense decoder transformers

The GPT design: one stack of transformer layers predicting the next token. 308 models on EveryLLM so far.

The dense decoder transformer is the design behind GPT and most chatbots. It is a single stack of identical transformer layers that reads the tokens so far and predicts the next one, with a mask that stops each position from looking ahead. Dense means every parameter is used for every token, unlike in a mixture of experts. GPT-3, PaLM, Llama 3, Chinchilla, Gemma, Phi and OLMo are dense decoders.

Inside each layer

Each layer has two parts. Self-attention lets every token weigh every earlier token and gather what it needs; a feed-forward network then transforms each token's vector on its own. The transformer was introduced by Google researchers in the June 2017 paper Attention Is All You Need, for translation, and OpenAI's GPT-1 (2018) kept only its decoder half. Llama 3.1 405B, released in July 2024, stacks 126 such layers.

Why it won

A decoder trains on any text without labels, because the next token is always the answer to predict, and it scales smoothly: scaling laws published by OpenAI in 2020 and DeepMind in 2022 showed the error falling predictably with more parameters, data and compute. Its weak point is cost. Attention grows with the square of the context length, and every parameter must be read for every token, which is why mixtures of experts became popular.

Famous dense decoder models

Still to be written: OLMo (language model).

Further reading

OLMo 2 🇺🇸 United StatesOLMo 2 Open language model family with full training transparencyGeneral · Open weights · 2024 Qwen-72B 🇨🇳 ChinaQwen-72B A 72-billion parameter open-weight language model from Alibaba Cloud.General · Open weights · 2023 imgMolmo 🇺🇸 United StatesMolmo Open vision-language model family by Allen Institute for AIMultimodal · Fully open · 2024 Qwen1.5 🇨🇳 ChinaQwen1.5 Open-weight language model series with nine sizes and strong multilingual support.General · Open weights · 2024 Pharia-1-LLM-7B 🇩🇪 GermanyPharia-1-LLM-7B Multilingual 7B parameter decoder model by Aleph Alpha ResearchGeneral · Open weights Poro 2 🇫🇮 FinlandPoro 2 Finnish and English instruction-following model based on Llama 3.1General · Conditional licence PhoGPT 🇻🇳 VietnamPhoGPT Open-source generative model for Vietnamese language tasks.General · Open weights · 2023 Lucie-7B 🇫🇷 FranceLucie-7B Open multilingual language model focused on French and EnglishGeneral · Open weights · 2025 CroissantLLM 🇫🇷 FranceCroissantLLM A 1.3 billion parameter bilingual French-English language model.General · Open weights · 2024 Bielik 🇵🇱 PolandBielik Polish-optimised 11 billion parameter language modelGeneral · Open weights · 2025 BLOOMZ WorldBLOOMZ Multilingual instruction-following model based on BLOOMGeneral · Open weights · 2022 fnQwQ-32B 🇨🇳 ChinaQwQ-32B A 32-billion parameter reasoning model from the Qwen series.Reasoning · Open weights · 2025 PLaMo-100B 🇯🇵 JapanPLaMo-100B A 100-billion parameter language model built for Japanese.General · Open weights · 2024 SeaLLMs 🇸🇬 SingaporeSeaLLMs Open language models for Southeast Asian languagesGeneral · Open weights · 2023 EvoLLM-JP 🇯🇵 JapanEvoLLM-JP Experimental Japanese language model created by merging open-source models.General · Open weights · 2024 fnMotif-2 WorldMotif-2 A 12.7 billion parameter reasoning model by Motif Technologies.Reasoning · Open weights MaLLaM 🇲🇾 MalaysiaMaLLaM Large language models trained on Malaysian context and Malay language data.General · Open weights · 2024 Taiwan-LLM 🇹🇼 TaiwanTaiwan-LLM Open language model for Traditional Chinese in TaiwanRegional · Fully open · 2023 YaLM 100B 🇷🇺 RussiaYaLM 100B Open-weight Russian and English language model with 100 billion parametersGeneral · Open weights HyperCLOVA X 🇰🇷 South KoreaHyperCLOVA X Korean-focused large language model with English and coding skillsGeneral · Research only · 2024 EXAONE 3.0 🇰🇷 South KoreaEXAONE 3.0 Open bilingual language model from LG AI ResearchGeneral · Open weights · 2024 Tsuzumi 🇯🇵 JapanTsuzumi Lightweight Japanese LLM for business useGeneral · API only · 2024 EXAONE 3.5 🇰🇷 South KoreaEXAONE 3.5 LG AI Research open language model series for real-world tasksGeneral · Conditional licence · 2024 fnEXAONE 4.0 🇰🇷 South KoreaEXAONE 4.0 Unified language model with reasoning and non-reasoning modesGeneral · Conditional licence · 2025 PanGu-Alpha 🇨🇳 ChinaPanGu-Alpha A large Chinese language model with 200 billion parameters.General · Research only · 2021 Kanana 🇰🇷 South KoreaKanana Bilingual language models for Korean and English by Kakao.General · Open weights · 2025 EuroLLM WorldEuroLLM Open multilingual models for European languagesGeneral · Open weights · 2024 Falcon-40B 🇦🇪 United Arab EmiratesFalcon-40B A 40-billion parameter open language model from TII.General · Open weights · 2023 LLM-jp-3 🇯🇵 JapanLLM-jp-3 Open Japanese language model family by NIIGeneral · Open weights · 2024 Llama-3-ELYZA-JP-8B 🇯🇵 JapanLlama-3-ELYZA-JP-8B Japanese-tuned Llama 3 model by ELYZAGeneral · Conditional licence Stockmark-100b 🇯🇵 JapanStockmark-100b A 100 billion parameter Japanese and English language model.General · Open weights MiniCPM 🇨🇳 ChinaMiniCPM Efficient small language models for end-side devicesGeneral · Open weights · 2024 Krutrim-1 🇮🇳 IndiaKrutrim-1 Multilingual language model for Indian languagesGeneral · Conditional licence · 2024 Krutrim-2 🇮🇳 IndiaKrutrim-2 A 12-billion parameter multilingual language model focused on Indian languages and context.General · Open weights · 2024 ERNIE 3.0 Titan 🇨🇳 ChinaERNIE 3.0 Titan Large-scale Chinese knowledge-enhanced pre-trained language modelGeneral · Research only · 2021 fnSeed-OSS-36B 🇨🇳 ChinaSeed-OSS-36B Open-source 36B parameter language model with controllable reasoning.General · Open weights · 2025 Pangu Ultra 🇨🇳 ChinaPangu Ultra A 135-billion parameter dense language model trained on Ascend hardware.General · API only · 2025 imgvidfnInternVL3 🇨🇳 ChinaInternVL3 Open-source multimodal large language model seriesMultimodal · Open weights · 2025 Baichuan 2 🇨🇳 ChinaBaichuan 2 Open multilingual language models with 7B and 13B parameters.General · Open weights · 2023 imgNVLM 1.0 🇺🇸 United StatesNVLM 1.0 Open multimodal language models from NVIDIA for vision and language tasksMultimodal · Open weights · 2024 K2-Think WorldK2-Think A 32 billion parameter open reasoning modelReasoning · Open weights · 2025 Fugaku-LLM 🇯🇵 JapanFugaku-LLM Japanese large language model trained on Fugaku supercomputerGeneral · Open weights · 2024 Qwen2.5-72B 🇨🇳 ChinaQwen2.5-72B A 72-billion parameter open-weight language model from Alibaba Cloud.General · Open weights · 2024 CodeGen 🇺🇸 United StatesCodeGen Open autoregressive language model for code generationCode · Open weights · 2022 EXAONE Deep 🇰🇷 South KoreaEXAONE Deep Reasoning-focused language models from LG AI ResearchReasoning · Open weights · 2025 DeepSeek Coder 🇨🇳 ChinaDeepSeek Coder Open-source code generation models from DeepSeekCode · Open weights · 2024 fnMistral Small 3 🇫🇷 FranceMistral Small 3 A 24B open-weight model tuned for low latency and local deploymentGeneral · Open weights · 2025 OPT 🇺🇸 United StatesOPT Meta AI's family of open decoder-only language models from 125M to 175B parametersGeneral · Conditional licence · 2022

The weekly digest

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