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

imgLlama 🇺🇸 United StatesLlama Meta AI's family of open-weight language models, from 1B to 2T parametersGeneral · Conditional licence · 2023 imgwavvidVideoPoet 🇺🇸 United StatesVideoPoet A Google Research language model that turns text and images into video and audioMultimodal · Research only · 2023 fnSparrow 🇬🇧 United KingdomSparrow DeepMind's rule-following English chatbot that cites web evidence for its answersAssistant · Research only · 2022 PaLM 🇺🇸 United StatesPaLM Google AI's 540-billion-parameter dense decoder-only transformer language modelGeneral · API only · 2022 LaMDA WorldLaMDA Google's conversational language model family, first shown in 2021General · API only · 2020 BuddhaBot 🇯🇵 Japan and 1 moreBuddhaBot A chatbot answering questions from early Buddhist textsDomain · Research only · 2021 Chinchilla 🇬🇧 United KingdomChinchilla A 70-billion-parameter DeepMind model built to test how models should scale.Research · Research only · 2022 imgfnGemini Robotics 🇺🇸 United StatesGemini Robotics A vision-language-action model from Google DeepMind built for robotsMultimodal · API only · 2025 Vicuna WorldVicuna An open chat model fine-tuned from LLaMA and Llama 2 by LMSYSGeneral · Conditional licence · 2023 1.58-bit large language model World1.58-bit large language model A model whose weights hold only three values: -1, 0 and +1.Research · Open weights · 2024 Stanford Alpaca 🇺🇸 United StatesStanford Alpaca A cheap instruction-tuned LLaMA 7B built at Stanford in 2023Research · Research only · 2023 Claude Mythos 🇺🇸 United StatesClaude Mythos Anthropic's restricted model family for finding software security flawsDomain · Research only · 2026 DialoGPT 🇺🇸 United StatesDialoGPT A GPT-style model trained on Reddit threads to write chat repliesGeneral · Open weights · 2019 Megatron-Turing NLG 530B 🇺🇸 United StatesMegatron-Turing NLG 530B A 530-billion-parameter generative language model trained by Microsoft and NVIDIA.General · Research only · 2022 Turing-NLG 🇺🇸 United StatesTuring-NLG Microsoft's 17-billion-parameter Transformer model for generation, QA and summarisationGeneral · Research only · 2020 BloombergGPT 🇺🇸 United StatesBloombergGPT A 50 billion parameter language model for finance.Domain · Research only · 2023 HyperCLOVA X 🇰🇷 South KoreaHyperCLOVA X Korean-focused large language model with English and coding skillsGeneral · Research only · 2024 EXAONE 3.5 🇰🇷 South KoreaEXAONE 3.5 LG AI Research open language model series for real-world tasksGeneral · Conditional licence · 2024 PanGu-Alpha 🇨🇳 ChinaPanGu-Alpha A large Chinese language model with 200 billion parameters.General · Research only · 2021 ERNIE 3.0 Titan 🇨🇳 ChinaERNIE 3.0 Titan Large-scale Chinese knowledge-enhanced pre-trained language modelGeneral · Research only · 2021 imgfnUI-TARS 🇨🇳 ChinaUI-TARS Open-source multimodal agent for GUI tasksMultimodal · Open weights · 2025 OPT 🇺🇸 United StatesOPT Meta AI's family of open decoder-only language models from 125M to 175B parametersGeneral · Conditional licence · 2022 AlphaCode 🇬🇧 United KingdomAlphaCode A code generation system that solves competitive programming problems.Code · Research only · 2022 RETRO WorldRETRO A retrieval-enhanced transformer that conditions on chunks from a huge document database.Research · Research only · 2021 GPT-SW3 🇸🇪 SwedenGPT-SW3 Generative language model for Nordic languagesGeneral · Research only · 2023 fnToolformer 🇺🇸 United StatesToolformer A language model that learns on its own when to call external tools through APIs.Research · Research only · 2023 fnNorth Mini Code 🇺🇸 United StatesNorth Mini Code A 30B sparse mixture-of-experts model for code and terminal agent workCode · Conditional licence HyperCLOVA 🇰🇷 South KoreaHyperCLOVA Korean language model with 82 billion parametersGeneral · Research only · 2021 imgMM1 WorldMM1 A family of multimodal language models up to 30B parameters, dense and MoE.Multimodal · Research only · 2024 Orca 2 🇺🇸 United StatesOrca 2 A small LLaMA-2-based model trained on synthetic data to reason.Small · Conditional licence · 2023 fnCommand A WorldCommand A Cohere's enterprise-focused large language model for agents, search and tool useGeneral · Research only · 2025 fnWebGPT 🇺🇸 United StatesWebGPT A GPT-3 model fine-tuned to answer questions while browsing the webGeneral · Research only · 2021 imgwavvidGemma 3n 🇺🇸 United StatesGemma 3n Mobile-first open model family for on-device text, image and audio workMultimodal · Open weights · 2025 Med-PaLM 🇺🇸 United StatesMed-PaLM A medical-domain language model aligned from Flan-PaLM using instruction prompt tuning.Domain · Research only · 2022 Med-PaLM 2 🇺🇸 United StatesMed-PaLM 2 Google's medical question-answering model built on PaLM 2Domain · Research only · 2023 Minerva 🇺🇸 United StatesMinerva A language model trained further on technical content to solve maths and science problemsResearch · Research only · 2022 Gopher 🇬🇧 United KingdomGopher A 280 billion parameter Transformer language model from DeepMind, studied for how scale changes ability.Research · Research only · 2021 imgChameleon WorldChameleon Early-fusion models that read and write images and text in one streamMultimodal · Research only · 2024 imgKimi k1.5 🇨🇳 ChinaKimi k1.5 Multimodal LLM trained with reinforcement learning for reasoningReasoning · Research only · 2025

The weekly digest

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