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

Talkie 🇺🇸 United StatesTalkie A 13B small language model trained only on pre-1931 public domain English textSmall · Open weights · 2026 imgPhi 🇺🇸 United StatesPhi Microsoft's small open-weights language models that run on a single deviceSmall · Open weights · 2023 imgwavvidGemma 3n 🇺🇸 United StatesGemma 3n Mobile-first open model family for on-device text, image and audio workMultimodal · Open weights · 2025 imgLlama 3.2 🇺🇸 United StatesLlama 3.2 Small text models and vision models from Meta for edge devices and image tasksMultimodal · Open weights · 2024 Phi-1 🇺🇸 United StatesPhi-1 A small Microsoft model specialised for basic Python codingCode · Open weights · 2023 imgPhi-3 🇺🇸 United StatesPhi-3 Small Microsoft language models built to run well on phones and laptops.Small · Open weights · 2024 Phi-4-mini 🇺🇸 United StatesPhi-4-mini A 3.8B-parameter compact language model from Microsoft for reasoning, maths and codeSmall · 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 imgMiniCPM-V 🇨🇳 ChinaMiniCPM-V Efficient multimodal model for mobile devicesMultimodal · Open weights · 2024 SOLAR 10.7B 🇰🇷 South KoreaSOLAR 10.7B An 11-billion parameter language model using depth up-scaling.General · Open weights · 2023 ChatGLM-6B 🇨🇳 ChinaChatGLM-6B Open bilingual dialogue model with 6.2 billion parametersAssistant · Open weights fnDevstral 🇫🇷 FranceDevstral A 24B agentic coding model built by Mistral AI and All Hands AICode · Open weights · 2025 OpenCoder WorldOpenCoder Open code model family with full training data and pipelineCode · Open weights · 2024 TinyLlama WorldTinyLlama A compact 1.1 billion parameter open language model.Small · Open weights · 2023 StableLM 2 🇬🇧 United KingdomStableLM 2 A small open-weight language model by Stability AIGeneral · Open weights · 2024 SmolLM2 🇺🇸 United StatesSmolLM2 Compact language models for on-device useGeneral · Open weights · 2025 Pharia-1-LLM-7B 🇩🇪 GermanyPharia-1-LLM-7B Multilingual 7B parameter decoder model by Aleph Alpha ResearchGeneral · Open weights Fathom-R1-14B WorldFathom-R1-14B Low-cost open reasoning model for mathematicsReasoning · Open weights MiniCPM 🇨🇳 ChinaMiniCPM Efficient small language models for end-side devicesGeneral · Open weights · 2024 Sarvam 2B 🇮🇳 IndiaSarvam 2B Small language model for English and ten Indian languages.Small · Open weights · 2024 imgfnDevstral 2 🇫🇷 FranceDevstral 2 Open-weight coding model family for software agentsCode · Open weights · 2026 OpenELM 🇺🇸 United StatesOpenELM Apple's small open language models with a full training and inference releaseSmall · Open weights · 2024 Pragna-1B 🇮🇳 IndiaPragna-1B Small decoder model for Hindi, Bangla, Gujarati and EnglishSmall · Open weights Granite 3.0 🇺🇸 United StatesGranite 3.0 IBM's third-generation open enterprise language models, from 2B to 8B parametersGeneral · Open weights · 2024 Mi:dm 2.0 🇰🇷 South KoreaMi:dm 2.0 Korea-centric language model by KTGeneral · Open weights · 2025 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

The weekly digest

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