Labsby GlyphNewest labs
founded most recentlyInstitut national pour l'évaluation et la sécurité de l'intelligence artificielle (INESIA)France's National Institute for AI Evaluation and Security, created by the French government on 31 January 2025.France
Meta Superintelligence LabsMeta's frontier model effort, formed in 2025 under Alexandr Wang; it builds the Muse models and now houses FAIR.Menlo Park, United StatesMuse Spark 1.1
Thinking Machines LabMira Murati's lab: Tinker, a fine-tuning API, and open research notes on reproducible inference.San Francisco, United States
Black Forest LabsThe FLUX image models, from the researchers behind latent diffusion and Stable Diffusion.Freiburg, GermanyFLUX
Canadian Artificial Intelligence Safety Institute (CAISI)Canada's AI safety institute, founded in November 2024 and led by ISED with NRC and CIFAR research partners.Canada
European AI OfficeEuropean Commission body that implements the EU AI Act, especially rules for general-purpose AI models.Brussels, Belgium
Japan AI Safety Institute (J-AISI)Japan's AI safety institute, launched on 14 February 2024 within IPA to develop AI safety evaluation methods and standards.Japan
Kling AI (Kuaishou)Kuaishou's generative video unit behind the Kling video model, plus open work such as LivePortrait and Kolors.Beijing, China
Korea AI Safety InstituteSouth Korea's AI safety institute, founded in November 2024 under ETRI and based in Bundang, Seongnam.Seongnam, South Korea
Microsoft AICopilot for consumers and Microsoft's own MAI models, led by Mustafa Suleyman since 2024.Redmond, United States
Physical IntelligenceGeneral-purpose robot foundation models: the pi-zero vision-language-action family, open-sourced as openpi.San Francisco, United States
Reflection AILab founded by former Google DeepMind researchers Misha Laskin and Ioannis Antonoglou to build open frontier models.Brooklyn, United States
Safe Superintelligence (SSI)Ilya Sutskever's lab with one goal and one product, a safe superintelligence; it has published nothing yet.Palo Alto, United States
01.AIKai-Fu Lee's lab behind the open Yi models, now focused on enterprise AI products.Beijing, ChinaYi-34B
Alibaba QwenThe Qwen family, the most downloaded open-weight models, from Alibaba Cloud's Tongyi lab.Hangzhou, ChinaQwen3.8 Max (0902)
Answer.AIR&D lab from Jeremy Howard and Eric Ries turning foundational research into practical tools, such as FSDP+QLoRA and ModernBERT.
Apollo ResearchAI safety organization focused on evaluating and monitoring scheming and deception in frontier AI systems.
Baichuan AIBeijing AI company, one of China's AI Tigers, known for the open Baichuan language models and medical models.Beijing, China
Contextual AIEnterprise RAG company founded by RAG co-author Douwe Kiela, known for RAG 2.0 and the KTO alignment method.Mountain View, United States
Databricks Mosaic ResearchOpen models (MPT, DBRX) and training at scale, from the MosaicML team Databricks bought in 2023; Spark lineage.San Francisco, United States
DatologyAIData curation company that researches how selecting and synthesizing training data makes models better and cheaper.Redwood City, United States
DecartReal-time generative video lab, known for Oasis, a playable AI-generated game world model.Tel Aviv, Israel
DeepSeekFrontier-class open-weight models trained cheaply: DeepSeek-V3's efficient MoE and R1's open reasoning recipe.Hangzhou, ChinaDeepSeek V4P1 Flash
H CompanyParis startup building computer-use agents and the open Holo models that navigate user interfaces.Paris, France
HarmonicCompany building mathematical superintelligence; its Aristotle system reached gold-medal level at the 2025 IMO.
Liquid AILiquid foundation models (LFM) for devices, built on liquid neural network research from MIT CSAIL.Cambridge, Massachusetts, United States
Mistral AIEurope's leading model maker: efficient open-weight models, Mixtral's mixture of experts, and Le Chat.Paris, FranceMistral Medium 3
Moonshot AIThe Kimi models: long context early, then the open-weight Kimi K2 trained with the Muon optimizer.Beijing, ChinaKimi K3
Sakana AINature-inspired methods: evolutionary model merging and The AI Scientist, from a Transformer co-author.Tokyo, Japan
Sarvam AIBengaluru company building language models for Indian languages, part of India's sovereign LLM programme.Bengaluru, India
SB IntuitionsSoftBank subsidiary developing homegrown Japanese large language models under the Sarashina name.Tokyo, Japan
Swiss AI InitiativeSwiss open foundation model effort started in December 2023 by ETH Zurich, EPFL and CSCS, which released the Apertus LLM.Switzerland
Tencent Hy (Hunyuan)Tencent's Hunyuan models, with open releases in MoE language, video (HunyuanVideo) and 3D generation.Shenzhen, China
UK AI Security InstituteThe first state AI safety institute (renamed Security in 2025): pre-deployment testing and the Inspect eval framework.United Kingdom
US Center for AI Standards and Innovation (CAISI)NIST's AI testing center, founded as the US AI Safety Institute and renamed CAISI in 2025; evaluates frontier and foreign models.United States
xAI (SpaceXAI)The Grok models and the Colossus supercomputer in Memphis; now a division of SpaceX.San Francisco Bay Area, United States
Center for AI Safety (CAIS)San Francisco nonprofit doing technical AI safety research and field-building; organized the 2023 Statement on AI Risk.San Francisco, United States
Character.AIConsumer chatbot company founded by LaMDA developers Noam Shazeer and Daniel de Freitas, training its own conversational models.United States
ElevenLabsAI audio research and product company known for natural-sounding speech synthesis and voice agent models.London, United Kingdom
Epoch AIData on AI's trajectory: training compute trends, data limits, and the FrontierMath benchmark.United States
Figure AIHumanoid robot company founded by Brett Adcock, developer of the Figure robots and the Helix vision-language-action model.Sunnyvale, United States
Inflection AIPublic benefit corporation that built Pi, a personal AI assistant powered by the Inflection AI model stack.Palo Alto, United States
METREvaluating dangerous capabilities: the time-horizon measure of how long a task AI agents can finish.Berkeley, California, United States
PerplexityAI answer engine company that also releases open models such as the pplx-embed embedding family.San Francisco, United States
RekaNatively multimodal models (Core, Flash, Edge) from former DeepMind and Meta researchers.Sunnyvale, United States
Singapore AI Safety Institute (Digital Trust Centre)Singapore's AI safety institute: the NTU Digital Trust Centre, founded in 2022 and designated the AISI in 2024, working with IMDA.Singapore
SunoMusic generation company whose models create songs with vocals from prompts; earlier released the open Bark audio model.Cambridge, Massachusetts, United States
Alignment Research Center (ARC)Berkeley non-profit founded by Paul Christiano in 2021, working on theory for mechanistic explanations of neural networks.Berkeley, United States
AnthropicThe Claude models, Constitutional AI and interpretability research on what happens inside models.San Francisco, United StatesClaude Fable 5.1
Distributed AI Research Institute (DAIR)Independent, community-rooted AI research institute founded by Timnit Gebru in December 2021.
ImbueAI lab founded in 2021 that trained a 70B model from scratch and builds tools that put software under users' control.United States
Isomorphic LabsAI-first drug discovery spun out of DeepMind; co-developed AlphaFold 3.London, United Kingdom
Kempner Institute for the Study of Natural and Artificial IntelligenceHarvard institute studying intelligence in brains and machines, funded by a Chan Zuckerberg gift in 2021.Allston, United States
MiniMaxLightning attention for million-token context in open models, plus Hailuo video and speech models.Shanghai, China
WaabiSelf-driving company founded by Raquel Urtasun, using generative AI simulation such as UniSim for autonomous trucks.Toronto, Canada
Beijing Institute for General Artificial Intelligence (BIGAI)Beijing AGI research institute founded in 2020 by Song-Chun Zhu with Beijing municipal government support.Beijing, China
EleutherAIGrassroots open research: The Pile, GPT-NeoX and Pythia, and the LM Evaluation Harness everyone uses.United States
ETH AI CenterETH Zurich's cross-department AI hub and a driver of the Swiss AI Initiative's open Apertus model.Zurich, Switzerland
Jina AISearch AI company founded by Han Xiao, known for open embedding, reranker and reader models.Berlin, Germany
Mohamed bin Zayed University of Artificial Intelligence (MBZUAI)Graduate research university dedicated to AI in Masdar City, Abu Dhabi, licensed in 2020.Abu Dhabi, United Arab Emirates
NSF AI Institute for Foundations of Machine Learning (IFML)NSF AI institute led by UT Austin, designated in 2020, working on foundational machine learning theory and tools.Austin, United States
Sony AISony Group's AI research subsidiary, known for the GT Sophy racing agent and the FHIBE fairness benchmark dataset.Japan
Technology Innovation Institute (TII)Abu Dhabi government-funded research institute, part of ATRC, that develops the Falcon open language models.Abu Dhabi, United Arab Emirates
UpstageSouth Korean AI company known for the Solar language models and depth up-scaling, plus document AI.Seoul, South Korea
Aleph AlphaGerman company building sovereign LLMs for enterprises and governments, including Pharia models and tokenizer-free T-FREE research.Heidelberg, Germany
Cohere and Cohere LabsEnterprise models (Command, Embed, Rerank) and Cohere Labs' open multilingual Aya research.Toronto, CanadaCommand A
Institute for Artificial Intelligence, Peking UniversityPeking University institute announced on 27 April 2019 to coordinate university-wide AI research and training.Beijing, China
Snorkel AIData-centric AI company spun out of the Stanford AI Lab, originating from the Snorkel weak supervision project.United States
Stability AIStable Diffusion, the open image model release that opened up generative imaging in 2022.London, United Kingdom
Stanford HAI and CRFMThe AI Index, the term foundation models, the HELM benchmark and Alpaca, from Stanford's AI institute.Stanford, California, United States
Z.ai (Zhipu AI)The GLM models, spun out of Tsinghua University's KEG lab; GLM-4.5 and later are open weights.Beijing, ChinaGLM-5.3
BAAI (Beijing Academy of AI)The BGE embedding models used across open retrieval stacks, the Emu multimodal models and FlagOpen.Beijing, ChinaBGE-M3
European Laboratory for Learning and Intelligent Systems (ELLIS)Pan-European AI network of excellence founded in 2018, with institutes and units at research institutions across Europe.
RunwayNew York generative video company behind the Gen-3 and Gen-4 video models and the GWM-1 world model.New York City, United States
Stanford DAWNFive-year Stanford project (2018 to 2022) on tools to democratize ML, including DAWNBench, Snorkel and ColBERT.Stanford, United States
Wadhwani Institute for Artificial Intelligence (Wadhwani AI)Independent non-profit founded in 2018 building AI for underserved communities in health, agriculture and education.Mumbai, India
AI Now InstituteIndependent research institute founded in 2017 producing policy research on the social impact of AI and industry power.New York, United States
AI SingaporeSingapore's national AI programme launched in May 2017, which develops the SEA-LION Southeast Asian language models.Singapore
AI21 LabsJamba, the first production-scale hybrid of Transformer and Mamba layers.Tel Aviv, Israel
DeepLGerman AI research company known for the DeepL Translator and its own translation-focused language models.Cologne, Germany
By country
1United States106
2China30
4Germany9
5Canada8
6Japan8
7France7
12Netherlands3
What changed
OpenAIFunding rounds updated.CMU Machine Learning DepartmentPapers on file rose from 9 to 10.NVIDIA ResearchPapers on file rose from 87 to 89.Alibaba QwenPapers on file rose from 74 to 78.Meta FAIRPapers on file rose from 39 to 40.Google ResearchPapers on file rose from 42 to 45.Scale AIRaised $1B (Series F) at a $13.8B valuation, May 2024.Berkeley Sky Computing LabPapers on file rose from 13 to 14.CMU Machine Learning DepartmentPapers on file rose from 8 to 9.Berkeley AI Research (BAIR)Papers on file rose from 13 to 14.NVIDIA ResearchPapers on file rose from 86 to 87.Alibaba QwenPapers on file rose from 71 to 74.Google ResearchPapers on file rose from 41 to 42.Microsoft ResearchPapers on file rose from 54 to 56.AnthropicPapers on file rose from 5 to 6.Alibaba QwenBest-ranked model is now Qwen3.8 Max (0902), was Qwen3.8 2.4T A95B.DeepSeekBest-ranked model is now DeepSeek V4P1 Flash.xAI (SpaceXAI)Grok 4.7 dropped out of the models.fru.dev ranking.Hugging FaceClément Delangue is now Chief Executive Officer and Co-founder.Databricks Mosaic ResearchAli Ghodsi is now Chief Executive Officer and Co-founder.Apple Machine Learning ResearchAmar Subramanya is now Vice President of AI.Safe Superintelligence (SSI)Daniel Gross is now Chief Executive Officer and Co-founder.Thinking Machines LabMira Murati is now Chief Executive Officer and Founder.Cohere and Cohere LabsAidan Gomez is now Chief Executive Officer and Co-founder.Mistral AIArthur Mensch is now Chief Executive Officer and Co-founder.Microsoft AIJordan Hoffmann is now Head of Microsoft AI London Hub.Meta Superintelligence LabsAlexandr Wang is now Chief AI Officer, Leads Meta Superintelligence Labs.Google DeepMindKoray Kavukcuoglu is now Chief AI Architect.AnthropicDario Amodei is now Chief Executive Officer and Co-founder.OpenAIBret Taylor is now Chair of the Board.CMU Machine Learning DepartmentPapers on file rose from 7 to 8.Stanford HAI and CRFMPapers on file rose from 9 to 12.Databricks Mosaic ResearchPapers on file rose from 1 to 3.NVIDIA ResearchPapers on file rose from 84 to 86.Alibaba QwenPapers on file rose from 65 to 71.Google ResearchPapers on file rose from 38 to 41.Microsoft ResearchPapers on file rose from 50 to 54.Baidu (ERNIE)Ernie 5.1 dropped out of the models.fru.dev ranking.Alibaba QwenBest-ranked model is now Qwen3.8 2.4T A95B, was Qwen3.8 Max (0902).DeepSeekDeepSeek V4.1 Flash dropped out of the models.fru.dev ranking.xAI (SpaceXAI)Best-ranked model is now Grok 4.7, was Grok 4.5.GradiumBuilds low-latency voice AI models for conversational agents; the $100M total includes a $30M extension of its seed round.Scale AIA round from scale.com no longer listed on Funding.NVIDIA ResearchPapers on file rose from 79 to 84.CMU Machine Learning DepartmentPapers on file rose from 6 to 7.Alibaba QwenPapers on file rose from 63 to 65.Microsoft ResearchPapers on file rose from 49 to 50.Meta FAIRPapers on file rose from 37 to 39.Google ResearchPapers on file rose from 36 to 38.Thinking Machines LabInkling dropped out of the models.fru.dev ranking.xAI (SpaceXAI)Best-ranked model is now Grok 4.5, was Grok 4.7.Meta Superintelligence LabsBest-ranked model is now Muse Spark 1.1, was Muse Spark.Google DeepMindBest-ranked model is now Gemini 4 Argon, was Gemini 3.8 Flash.AnthropicDurk Kingma and Diederik P. Kingma are one person (the same Wikidata item), now listed as Diederik P. Kingma.Nous ResearchRaised $90M (Series B) at a $1.5B valuation, Oct 2026.NVIDIA ResearchPapers on file rose from 75 to 79.CMU Machine Learning DepartmentPapers on file rose from 5 to 6.Berkeley Sky Computing LabPapers on file rose from 12 to 13.Berkeley AI Research (BAIR)Papers on file rose from 12 to 13.Alibaba QwenPapers on file rose from 61 to 63.Microsoft ResearchPapers on file rose from 47 to 49.Meta FAIRPapers on file rose from 34 to 37.Google ResearchPapers on file rose from 33 to 36.xAI (SpaceXAI)Best-ranked model is now Grok 4.7, was Grok 4 1 Fast.ElevenLabsRaised $300M at a $22B valuation, Sep 2026.CMU Machine Learning DepartmentPapers on file rose from 4 to 5.Microsoft ResearchPapers on file rose from 45 to 47.Meta Superintelligence LabsBest-ranked model is now Muse Spark, was Muse Spark 1.3.World LabsRaised $230M at a $1B valuation, Sep 2024.Physical IntelligenceRaised $400M (Series A) at a $2.4B valuation, Nov 2024.Stability AIA new round, Jun 2024.Black Forest LabsRaised $31M (Seed), Aug 2024.Safe Superintelligence (SSI)Raised $1B at a $5B valuation, Sep 2024.Thinking Machines LabRaised $2B (Seed) at a $12B valuation, Jul 2025.Moonshot AIRaised $2B (Series D) at a $20B valuation, May 2026.DeepSeekRaised $7.4B at a $50B valuation, Jun 2026.AnthropicRaised $2.8B (Strategic), Mar 2024.Mistral AIA new round (Series C), Sep 2025.OpenAIRaised $6.6B (Growth) at a $157B valuation, Oct 2024.Berkeley AI Research (BAIR)Papers on file rose from 2 to 12.