AllLLMRoboticsStartupsResearchEthicsBusinessHardwareAutonomousMedicineCreative
Topic hub

Meta AI

Meta AI is the AI arm of Meta (Facebook, Instagram, WhatsApp) and the driving force behind the open-weight Llama model family. Its open releases power thousands of derivative models, and its bets on smart glasses and superintelligence labs shape the industry's direction. All our Meta AI coverage lives on this page.

Official site ↗1064 articles
In the glossary:Large Language Model (LLM)Foundation Model
Fresh
NEWWiredWired Tested 6 AI Agents in a Day: Dots Leads, Gemini Spark Fails at LunchHugging Face BlogLiquid AI releases LFM2.5-24B-A2B: an agent model with 2B active parametersInterconnects (Nathan Lambert)Meta releases Muse Spark Open: 816B parameters and the lead among US open-weight modelsNVIDIA AI BlogNVIDIA Releases 120B Nemotron 3.5 Super VL and Bolt Models for Local AgentsLatent Space (swyx)Goodfire: Agentic Analysis Decodes Neurons Once Written Off as ChaosHabr AIGemma 26B fine-tuned on a 24 GB MacBook: offline assistant gets 703 of 766 rightHugging Face BlogDharma-AI trimmed Qwen3-4B's vocabulary and cut inference cost by 77%Hugging Face BlogMeta open-sources Muse Spark Open: an 816B-parameter MoE with a 1M-token context
Comparisons
Llama vs MistralDeepSeek vs Llama
Wired Tested 6 AI Agents in a Day: Dots Leads, Gemini Spark Fails at Lunch
LLM·Wired

Wired Tested 6 AI Agents in a Day: Dots Leads, Gemini Spark Fails at Lunch

Oct 28, 2026·3 min
Liquid AI releases LFM2.5-24B-A2B: an agent model with 2B active parameters
LLM

Liquid AI releases LFM2.5-24B-A2B: an agent model with 2B active parameters

Oct 28, 2026·3 min
Meta releases Muse Spark Open: 816B parameters and the lead among US open-weight models
LLM

Meta releases Muse Spark Open: 816B parameters and the lead among US open-weight models

Oct 28, 2026·3 min
NVIDIA Releases 120B Nemotron 3.5 Super VL and Bolt Models for Local Agents
LLM

NVIDIA Releases 120B Nemotron 3.5 Super VL and Bolt Models for Local Agents

Oct 28, 2026·3 min
Goodfire: Agentic Analysis Decodes Neurons Once Written Off as Chaos
LLM·Latent Space (swyx)

Goodfire: Agentic Analysis Decodes Neurons Once Written Off as Chaos

Goodfire's Nick Cammarata argues neural networks are "already clean": in 2023 only about 2% of neurons were considered readable, and many famous polysemanticity…

Oct 28, 2026·3 min
Gemma 26B fine-tuned on a 24 GB MacBook: offline assistant gets 703 of 766 right
LLM·Habr AI

Gemma 26B fine-tuned on a 24 GB MacBook: offline assistant gets 703 of 766 right

A developer fine-tuned Gemma 4 26B A4B with QLoRA on a laptop and moved 5 000 tokens of rules into the weights. Result: 703 of 766 conversations correct, median…

Oct 28, 2026·4 min
Dharma-AI trimmed Qwen3-4B's vocabulary and cut inference cost by 77%
LLM·Hugging Face Blog

Dharma-AI trimmed Qwen3-4B's vocabulary and cut inference cost by 77%

Brazil's Dharma-AI rebuilt the Qwen3-4B tokenizer for a single task. The average response shrank from 350 to 101 tokens, and the benchmark score rose from 0.703 to…

Oct 28, 2026·3 min
Meta open-sources Muse Spark Open: an 816B-parameter MoE with a 1M-token context
LLM·Hugging Face Blog

Meta open-sources Muse Spark Open: an 816B-parameter MoE with a 1M-token context

Meta has released Muse Spark Open, an open-weight model with 816B parameters, 42B active per token, text, image, audio and video input, and tuning for agentic…

Oct 28, 2026·3 min
Top themes
01Property Finder Found a Flaw in Its Labels and Cut VLM Costs by 86%Oct 28, 2026·4 min02DeepSeek Open-Sources DeepInfer as the US and China Prepare Controls on Open ModelsOct 28, 2026·3 min03Anthropic Files S-1: $11.6 Billion in Quarterly Revenue and $780 Billion in CommitmentsOct 28, 2026·3 min04Open-source Hob brings Claude Cowork-style workflows to your laptop, no cloud requiredOct 27, 2026·3 min05The $448 Billion Paradox: Why 95% of Corporate AI Pilots Deliver Zero P&L ImpactOct 27, 2026·3 min
DSReg Ties JEPA's Latent Variables to the Real Causes Behind the Data
LLM·Jiqizhixin (机器之心)

DSReg Ties JEPA's Latent Variables to the Real Causes Behind the Data

Oct 26, 2026·3 min
Alibaba Launches Second-Generation Qwen Glasses Starting at $450
LLM·vc.ru

Alibaba Launches Second-Generation Qwen Glasses Starting at $450

Oct 26, 2026·2 min
4.5x Faster Product Launches and 87% Budget Savings: 3 Secure AI Architectures Under Russia's 152-FZ
Opinion·Жемал Хамидун · Авторская колонка

4.5x Faster Product Launches and 87% Budget Savings: 3 Secure AI Architectures Under Russia's 152-FZ

Oct 26, 2026·3 min
Anthropic Replaces Git With Its Own System, Scion: 7 Engineers and Claude in 5 Months
LLM·Jiqizhixin (机器之心)

Anthropic Replaces Git With Its Own System, Scion: 7 Engineers and Claude in 5 Months

Oct 26, 2026·2 min
TechCrunch rounds up AI agents that live in text messages, from personal assistants to family and work helpers
LLM·TechCrunch

TechCrunch rounds up AI agents that live in text messages, from personal assistants to family and work helpers

TechCrunch compiled a list of AI agents that work directly inside text messages and sorted them into 4 types: general-purpose, family, travel and work.

Oct 26, 2026·2 min
OpenAI Unveils Dots Agent and Promises a 'New Privacy Standard' in AI
LLM·The Verge

OpenAI Unveils Dots Agent and Promises a 'New Privacy Standard' in AI

At DevDay, Altman introduced the Dots agent and promised a "new privacy standard." Months earlier, Zuckerberg said the same about Muse. Which of them will keep…

Oct 26, 2026·2 min
Instinct, August's Most Talked-About AI Agent, Now Faces Muse and Dots
LLM·The Verge

Instinct, August's Most Talked-About AI Agent, Now Faces Muse and Dots

The startup launched a text-based AI agent with no ads and barely a website, then got two Big Tech rivals within a couple of months. Is a first-mover advantage enough for Instinct?

Oct 26, 2026·3 min
Open AI Models: Democratization vs. Cyber Threats
LLM·MIT Technology Review

Open AI Models: Democratization vs. Cyber Threats

Oct 25, 2026·3 min
Lovable Teaches Its AI to Promote Apps: Google and Meta Ads From the Chat
LLM·Lovable Blog

Lovable Teaches Its AI to Promote Apps: Google and Meta Ads From the Chat

Oct 25, 2026·3 min
Meta releases its first frontier models: Muse Wonder and Muse Wonder Full
LLM·TechCrunch

Meta releases its first frontier models: Muse Wonder and Muse Wonder Full

Oct 25, 2026·2 min
Big Tech Goes Into Debt: Alphabet and Meta Borrow Billions for AI
LLM·Bloomberg Tech

Big Tech Goes Into Debt: Alphabet and Meta Borrow Billions for AI

Oct 25, 2026·2 min
a16z measures AI spending for the first time: only 4.5% of Americans pay, top 1% spends $900
LLM·The Decoder

a16z measures AI spending for the first time: only 4.5% of Americans pay, top 1% spends $900

Oct 25, 2026·2 min
Microsoft Prepares to Install Gigawatts of Its Own Chips in Data Centers
LLM

Microsoft Prepares to Install Gigawatts of Its Own Chips in Data Centers

Oct 25, 2026·2 min
Meta releases Muse Wonder and Wonder Full: intelligence index rises from 48 to 56
LLM

Meta releases Muse Wonder and Wonder Full: intelligence index rises from 48 to 56

Oct 24, 2026·3 min
Artificial Analysis launches AA-3D: Claude Sonnet 5.5 leads with 1194 Elo
LLM

Artificial Analysis launches AA-3D: Claude Sonnet 5.5 leads with 1194 Elo

Oct 24, 2026·3 min
State of IT in 2026: Agents Wrote More Code Than Humans
LLM·Habr AI

State of IT in 2026: Agents Wrote More Code Than Humans

In August 2026, agents opened more pull requests on GitHub than humans did. Since July, they've filed more tickets on Linear too. IDEs are disappearing, every…

Oct 24, 2026·3 min
Nvidia in talks to buy Reflection AI, a startup valued at $8 billion
LLM·Hacker News AI

Nvidia in talks to buy Reflection AI, a startup valued at $8 billion

Financial Times reports Nvidia is in talks to acquire Reflection AI, a US open-model developer valued at $8 billion in 2025 after a $2 billion round.

Oct 24, 2026·2 min