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

Apertus 🇨🇭 SwitzerlandApertus A Swiss public large language model released in 8B and 70B sizesGeneral · Open weights · 2025 imgwavvidQwen 🇨🇳 ChinaQwen A family of large and small language models developed by Alibaba Cloud.General · Open weights · 2023 Jais 🇦🇪 United Arab EmiratesJais An open Arabic and English large language model from the UAEGeneral · Open weights · 2023 Velvet AI 🇮🇹 ItalyVelvet AI Italian multilingual open-weight LLMs built for energy efficiencyGeneral · Open weights · 2025 GPT-2 🇺🇸 United StatesGPT-2 OpenAI's 2019 text model that predicts the next word in a sequenceGeneral · Open weights · 2019 GPT-1 WorldGPT-1 OpenAI's first GPT model, a 117M-parameter decoder-only transformer from 2018.Research · Open weights · 2018 IBM Granite 🇺🇸 United StatesIBM Granite IBM's decoder-only foundation models, from 3 to 34 billion parametersGeneral · Open weights · 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 Minerva 🇮🇹 ItalyMinerva An Italian-first large language model trained from scratch at Sapienza University of Rome.General · Open weights · 2024 XLNet WorldXLNet Permutation-based autoregressive transformer model created by Google AIResearch · Open weights · 2019 imgPhi 🇺🇸 United StatesPhi Microsoft's small open-weights language models that run on a single deviceSmall · Open weights · 2023 imgwavvidGemma 🇺🇸 United StatesGemma Google DeepMind's open family of Gemini-derived language modelsGeneral · Open weights · 2024 GPT-J WorldGPT-J A 6-billion-parameter open English text generator from EleutherAI, released in 2021.General · Open weights · 2021 GPT-4chan WorldGPT-4chan A GPT-J model fine-tuned on posts from 4chan's /pol/ board.General · Open weights · 2022 fnGLM 🇨🇳 ChinaGLM Open-weight Chinese model series used for chat and codingGeneral · Open weights · 2021 Talkie 🇺🇸 United StatesTalkie A 13B small language model trained only on pre-1931 public domain English textSmall · Open weights · 2026 imgHuawei PanGu 🇨🇳 ChinaHuawei PanGu Huawei's multimodal large model family for industry, from 1B to trillion-scaleMultimodal · Open weights · 2023 imgwavvidGemma 4 🇺🇸 United StatesGemma 4 Google DeepMind's open multimodal models in five sizes, from phones to serversGeneral · Open weights imgMiniCPM-V 🇨🇳 ChinaMiniCPM-V Efficient multimodal model for mobile devicesMultimodal · Open weights · 2024 imgwavvidGemma 3n 🇺🇸 United StatesGemma 3n Mobile-first open model family for on-device text, image and audio workMultimodal · Open weights · 2025 imgvidfnGemma 3 🇺🇸 United StatesGemma 3 Lightweight multimodal open models from 1B to 27B parametersGeneral · Open weights · 2025 Phi-4-reasoning 🇺🇸 United StatesPhi-4-reasoning A 14B open-weight reasoning model fine-tuned from Phi-4Reasoning · Open weights · 2025 imgwavPhi-4-multimodal 🇺🇸 United StatesPhi-4-multimodal A small open model that reads text, images and audio and replies in textMultimodal · Open weights · 2025 Gemma 2 🇺🇸 United StatesGemma 2 Open lightweight language models from Google DeepMind, 2B to 27B parametersGeneral · Open weights · 2024 Falcon2-11B 🇦🇪 United Arab EmiratesFalcon2-11B An 11 billion parameter causal decoder model by TII.General · Open weights · 2024 Falcon 3 🇦🇪 United Arab EmiratesFalcon 3 Open foundation models from 1B to 10B parameters by TIIGeneral · Open weights · 2024 Falcon-7B 🇦🇪 United Arab EmiratesFalcon-7B Open 7-billion parameter decoder model by TIIGeneral · Open weights · 2023 Sarvam-M 🇮🇳 IndiaSarvam-M Hybrid reasoning model for Indian languages and EnglishGeneral · Open weights · 2025 imgYi-34B 🇨🇳 ChinaYi-34B A 34-billion parameter bilingual language model by 01.AIGeneral · Open weights · 2023 Falcon 180B 🇦🇪 United Arab EmiratesFalcon 180B Open 180 billion parameter language model from TIIGeneral · Open weights · 2023 Phi-4-mini 🇺🇸 United StatesPhi-4-mini A 3.8B-parameter compact language model from Microsoft for reasoning, maths and codeSmall · Open weights · 2025 ChatGLM-6B 🇨🇳 ChinaChatGLM-6B Open bilingual dialogue model with 6.2 billion parametersAssistant · Open weights imgPhi-3 🇺🇸 United StatesPhi-3 Small Microsoft language models built to run well on phones and laptops.Small · Open weights · 2024 Phi-1 🇺🇸 United StatesPhi-1 A small Microsoft model specialised for basic Python codingCode · Open weights · 2023 fnBlenderBot 3 WorldBlenderBot 3 A 175B open-domain chatbot with internet access and long-term memoryGeneral · Open weights · 2022 Phi-2 🇺🇸 United StatesPhi-2 A 2.7 billion parameter Microsoft research model for QA, chat and codeResearch · Open weights Nemotron-4-Mini-Hindi-4B-Instr 🇺🇸 United StatesNemotron-4-Mini-Hindi-4B-Instruct A small Hindi-focused chat model from NVIDIA, tuned for Indian-context answersSmall · Open weights · 2024 StarCoder2 WorldStarCoder2 Open code generation model by BigCodeCode · Open weights · 2024 Pythia 🇺🇸 United StatesPythia A suite of 16 models for interpretability research.Research · Open weights · 2023 PLLuM 🇵🇱 PolandPLLuM Polish-focused large language model familyGeneral · Open weights · 2024 ULMFiT WorldULMFiT Transfer learning method for NLP text classification tasksResearch · Open weights · 2018 INTELLECT-1 WorldINTELLECT-1 A 10 billion parameter model trained by a global community.General · Open weights · 2024 Llama Guard WorldLlama Guard A Llama 2 based model that classifies prompts and replies for safety risksDomain · Open weights · 2023 RakutenAI-7B 🇯🇵 JapanRakutenAI-7B Japanese-focused open language model based on MistralGeneral · Open weights · 2024 SOLAR 10.7B 🇰🇷 South KoreaSOLAR 10.7B An 11-billion parameter language model using depth up-scaling.General · Open weights · 2023 Phi-4 🇺🇸 United StatesPhi-4 A 14B Microsoft Research model trained heavily on synthetic data for reasoning.General · Open weights · 2024 Sarashina2 🇯🇵 JapanSarashina2 Open-weight Japanese and English language models by SB IntuitionsGeneral · Open weights SEA-LION 🇸🇬 SingaporeSEA-LION Multilingual language models for Southeast AsiaGeneral · Open weights · 2025

The weekly digest

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