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

imgfnGPT-5.3-Codex 🇺🇸 United StatesGPT-5.3-Codex OpenAI's code-focused large language model with repository and terminal accessCode · API only · 2026 fnOpenAI Codex 🇺🇸 United StatesOpenAI Codex A code-writing language model fine-tuned from GPT-3Code · API only · 2025 IBM Granite 🇺🇸 United StatesIBM Granite IBM's decoder-only foundation models, from 3 to 34 billion parametersGeneral · Open weights · 2023 fnGPT-5.6 🇺🇸 United StatesGPT-5.6 OpenAI's 2026 model family in Luna, Terra and Sol tiersGeneral · API only · 2026 fnGLM-4.7 🇨🇳 ChinaGLM-4.7 Coding and reasoning model from Z.aiCode · Open weights fnMiniMax-M2.5 🇨🇳 ChinaMiniMax-M2.5 Cost-efficient coding and agentic model from MiniMax AICode · API only · 2026 fnGrok Code Fast 1 🇺🇸 United StatesGrok Code Fast 1 A fast, low-cost reasoning model built for agentic coding work.Reasoning · API only · 2026 imgfnDevstral 2 🇫🇷 FranceDevstral 2 Open-weight coding model family for software agentsCode · Open weights · 2026 imgClaude Opus 4.7 🇺🇸 United StatesClaude Opus 4.7 Anthropic's large model for long-running coding and professional workGeneral · API only · 2026 fnMiniMax-M2.7 🇨🇳 ChinaMiniMax-M2.7 Open-weight model for coding and agent tasksCode · Open weights DeepSeek-V2 🇨🇳 ChinaDeepSeek-V2 Efficient mixture-of-experts language model with 236 billion parametersGeneral · Open weights · 2024 fnClaude Haiku 4.5 🇺🇸 United StatesClaude Haiku 4.5 Anthropic's small, fast Claude model for coding, agents and real-time workGeneral · API only · 2025 fnCodestral 🇫🇷 FranceCodestral Mistral AI's 22B open-weight model built for code generation across 80+ languagesCode · Conditional licence fnDevstral 🇫🇷 FranceDevstral A 24B agentic coding model built by Mistral AI and All Hands AICode · Open weights · 2025 fnDevin 🇺🇸 United StatesDevin Autonomous software engineering agent by Cognition LabsCode · API only

The weekly digest

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