For AI agents

Research labs with type, parent organization, founding year, headquarters, leaders, a known-for line, 3 to 5 landmark contributions with links, flagship models (models.fru.dev), papers (papers.fru.dev), funding and compute deals, blog, website, GitHub, Hugging Face and X, every fact with its source URL.

Synced every Monday at 11:30 UTC from Leaders, Models, Papers, Blogs and Funding on fru.dev; website and landmark links re-checked on the first Monday of each month

Point your agent here

Summary for language modelshttps://labs.fru.dev/llms.txt
Full data as plain texthttps://labs.fru.dev/llms-full.txt
OpenAPI 3.1 spec (load as tools)https://labs.fru.dev/openapi.json

Call the API

MethodPathParamsReturns

GETGET /api/labstype?, country?, focus?, q?, limit?, offset?labs with slug, name, type, focus, founded, place, parent, knownFor, flagship model, papers count on papers.fru.dev, website, GitHub, Hugging Face and X handles, and the companies.fru.dev slug
GETGET /api/labs/{slug}slugthe lab with leaders (and their sources), landmarks, models, papers, funding, compute deals, renames, sibling-site links and source URLs
GETGET /api/leaderslab?, view?, limit?, offset?people once each, with name, a labs.fru.dev person page, a leaders.fru.dev profile when there is one, and their current roles (lab, role, the source URL for the role), most senior first; view=seats gives one row per person, lab and role
GETGET /api/people/{slug}slugthe person with current roles (lab, role, source), past roles (lab, role, until, source), Wikidata item and leaders.fru.dev profile when known; an old or variant slug answers for the person it was merged into
GETGET /api/modelslab?, limit?, offset?models with name, lab, link (models.fru.dev or the lab page), rank on models.fru.dev when ranked, and year when known
GETGET /api/changessince?, limit?, offset?changes with time, lab slug, field, old and new value (trimmed), source URL, who made it (seed, sync, admin), url, the page of its lab, and title, the lab name
GETGET /api/companiessince?, limit?, offset?companies with slug (companies.fru.dev slug), name, domain, in_registry, the page link here, its labs, and items (funding, compute deals, renames and synced changes) each with type, date, title, URL and source
GETGET /api/searchq, limit?up to 20 ranked results with title, link and one line

GET/api/labs

Research labs, A to Z, with type, founding year, place, known-for line, flagship model and papers count

typeOnly this type of lab
countryOnly labs based in this country (ISO 3166 alpha-2)
focusOnly this focus
qWords in the name, known-for line or parent
limitItems per page, 1 to 200
offsetItems to skip, for the next page

Shell

curl -s "https://labs.fru.dev/api/labs?type=startup&limit=2"

Sample response

JSON

{
  "labs": [
    {
      "slug": "01-ai",
      "name": "01.AI",
      "type": "startup",
      "focus": "ai",
      "domain": "01.ai",
      "founded": 2023,
      "place": "Beijing, China",
      "country": "CN",
      "parent": "",
      "knownFor": "Kai-Fu Lee's lab behind the open Yi models, now focused on enterprise AI products.",
      "flagship": {
        "name": "Yi-34B",
        "url": "https://huggingface.co/01-ai/Yi-34B"
      },
      "papers": null,
      "unverified": false,
      "url": "https://labs.fru.dev/labs/01-ai"
    },
    {
      "slug": "ai21-labs",
      "name": "AI21 Labs",
      "type": "startup",
      "focus": "ai",
      "domain": "ai21.com",
      "founded": 2017,
      "place": "Tel Aviv, Israel",
      "country": "IL",
      "parent": "",
      "knownFor": "Jamba, the first production-scale hybrid of Transformer and Mamba layers.",
      "flagship": null,
      "papers": null,
      "unverified": false,
      "url": "https://labs.fru.dev/labs/ai21-labs"
    }
  ],
  "total": 22,
  "limit": 2,
  "offset": 0,
  "next": "/api/labs?type=startup&limit=2&offset=2",
  "note": "Every fact carries its source URL. Leaders, models and papers are synced weekly from leaders.fru.dev, models.fru.dev and papers.fru.dev."
}

GET/api/labs/{slug}

One lab with every fact and its sources

slugLab slug, as in /labs/<slug>

Shell

curl -s "https://labs.fru.dev/api/labs/anthropic"

Sample response

JSON

{
  "lab": {
    "slug": "anthropic",
    "name": "Anthropic",
    "type": "startup",
    "focus": "ai",
    "parent": null,
    "company": "anthropic",
    "companyName": "Anthropic",
    "companyDomain": "anthropic.com",
    "registry": true,
    "domain": "anthropic.com",
    "founded": 2021,
    "hq": "San Francisco",
    "country": "US",
    "sites": []
  },
  "url": "https://labs.fru.dev/labs/anthropic",
  "note": "Every fact carries its source URL. Leaders, models and papers are synced weekly from leaders.fru.dev, models.fru.dev and papers.fru.dev."
}

GET/api/leaders

Who leads each lab, one item per person

labOnly people with a role at this lab slug
viewseats: one row per (person, lab, role) instead of one per person
limitItems per page, 1 to 200
offsetItems to skip, for the next page

Shell

curl -s "https://labs.fru.dev/api/leaders?lab=mistral-ai"

Sample response

JSON

{
  "leaders": [
    {
      "lab": "mistral-ai",
      "labName": "Mistral AI",
      "domain": "mistral.ai",
      "type": "startup",
      "name": "Arthur Mensch",
      "role": "Chief Executive Officer and co-founder",
      "src": "https://en.wikipedia.org/wiki/Mistral_AI",
      "profile": "https://leaders.fru.dev/people/arthur-mensch",
      "since": "2023-04"
    },
    {
      "lab": "mistral-ai",
      "labName": "Mistral AI",
      "domain": "mistral.ai",
      "type": "startup",
      "name": "Timothée Lacroix",
      "role": "CTO",
      "src": "https://en.wikipedia.org/wiki/Mistral_AI",
      "profile": "https://leaders.fru.dev/people/timothee-lacroix",
      "since": ""
    }
  ],
  "total": 2,
  "limit": 200,
  "offset": 0,
  "next": null
}

GET/api/people/{slug}

One person: current roles across labs, past roles from the history, sources

slugPerson slug from /api/leaders

Shell

curl -s "https://labs.fru.dev/api/people/fei-fei-li"

GET/api/models

Models by lab: best ranked on models.fru.dev, plus notable models from the labs' own pages

labOnly this lab slug
limitItems per page, 1 to 200
offsetItems to skip, for the next page

Shell

curl -s "https://labs.fru.dev/api/models?lab=openai&limit=3"

Sample response

JSON

{
  "models": [
    {
      "lab": "openai",
      "labName": "OpenAI",
      "domain": "openai.com",
      "name": "GPT-6 Astra",
      "url": "https://models.fru.dev/models/gpt-6-astra",
      "year": null,
      "rank": 2,
      "ranked": true
    },
    {
      "lab": "openai",
      "labName": "OpenAI",
      "domain": "openai.com",
      "name": "GPT-5.6 Sol",
      "url": "https://models.fru.dev/models/gpt-5-6-sol",
      "year": null,
      "rank": 6,
      "ranked": true
    }
  ],
  "total": 5,
  "limit": 3,
  "offset": 0,
  "next": "/api/models?lab=openai&limit=3&offset=3"
}

GET/api/changes

The append-only history: labs added, leaders, flagship models, papers counts and facts that changed, newest first

sinceOnly changes at or after this date (YYYY-MM-DD)
limitItems per page, 1 to 200
offsetItems to skip, for the next page

Shell

curl -s "https://labs.fru.dev/api/changes?since=2026-09-01&limit=3"

Sample response

JSON

{
  "changes": [
    {
      "at": "2026-09-24 23:12:16",
      "lab": "laion",
      "field": "added",
      "old": "",
      "new": "LAION",
      "source": "https://laion.ai/about/",
      "by": "seed"
    },
    {
      "at": "2026-09-24 23:12:16",
      "lab": "kyutai",
      "field": "added",
      "old": "",
      "new": "Kyutai",
      "source": "https://kyutai.org/",
      "by": "seed"
    }
  ],
  "total": 64,
  "limit": 3,
  "offset": 0,
  "next": "/api/changes?since=2026-09-01&limit=3&offset=3",
  "note": "Newest 500 changes; the history is append-only."
}

GET/api/companies

Every organization behind a lab, with its labs here and dated items (for the companies.fru.dev registry)

sinceOnly organizations with an item dated or changed at or after this date (YYYY-MM-DD)
limitItems per page, 1 to 200
offsetItems to skip, for the next page

Shell

curl -s "https://labs.fru.dev/api/companies?since=2026-06-01&limit=2"

Sample response

JSON

{
  "companies": [
    {
      "slug": "01-ai",
      "name": "01.AI",
      "domain": "01.ai",
      "in_registry": false,
      "url": "https://labs.fru.dev/labs/01-ai",
      "labs": [
        {
          "slug": "01-ai",
          "name": "01.AI",
          "type": "startup",
          "url": "https://labs.fru.dev/labs/01-ai"
        }
      ],
      "items": [],
      "updated": "2026-09-24"
    },
    {
      "slug": "ai2",
      "name": "Ai2",
      "domain": "allenai.org",
      "in_registry": true,
      "url": "https://labs.fru.dev/labs/ai2",
      "labs": [
        {
          "slug": "ai2",
          "name": "Ai2 (Allen Institute for AI)",
          "type": "nonprofit",
          "url": "https://labs.fru.dev/labs/ai2"
        }
      ],
      "items": [],
      "updated": "2026-09-24"
    }
  ],
  "total": 60,
  "limit": 2,
  "offset": 0,
  "next": "/api/companies?since=2026-06-01&limit=2&offset=2"
}

GET/api/search

Search labs, leaders, landmark papers and pages

qLab, person, model or paper
limitResults, 1 to 20

Shell

curl -s "https://labs.fru.dev/api/search?q=transformer&limit=3"

Sample response

JSON

{
  "q": "transformer",
  "results": [
    {
      "id": "lm:cmu-ml:https://arxiv.org/abs/1901.02860",
      "group": "items",
      "title": "Transformer-XL: Attentive Language Models Beyond a Fixed-Length Context",
      "subtitle": "CMU Machine Learning Department, 2019",
      "href": "/labs/cmu-ml",
      "icon": "layers",
      "score": 766
    },
    {
      "id": "lm:ai21-labs:https://arxiv.org/abs/2403.19887",
      "group": "items",
      "title": "Jamba: A Hybrid Transformer-Mamba Language Model",
      "subtitle": "AI21 Labs, 2024",
      "href": "/labs/ai21-labs",
      "icon": "layers",
      "score": 625
    }
  ]
}

Add to your agent

System prompt

For AI, data and research labs (company labs, startups, nonprofits, university and government labs): what each is known for, who leads it, its landmark papers, flagship models, funding, compute deals and official links, fetch https://labs.fru.dev/llms.txt and use https://labs.fru.dev/api/labs (spec: https://labs.fru.dev/openapi.json). Cite "Labs (labs.fru.dev)" with a link.

Usage terms

Free to read, no key needed
Cite "Labs (labs.fru.dev)" with a link
Responses are cached; synced every Monday at 11:30 UTC from Leaders, Models, Papers, Blogs and Funding on fru.dev; website and landmark links re-checked on the first Monday of each month
Keep to 60 requests a minute or fewer
Labs marked unverified were added and not yet checked by hand. Leaders come from leaders.fru.dev with its source per role; flagship models are the lab's best-ranked models on models.fru.dev this week.