Every API here is available over the APIs.io API and to AI agents over MCP.
MCP server
One button, every client — Claude, Cursor, VS Code and the rest.
https://apis.io/mcp
Tools for apis
7 MCP tools reach this
find_apisBrowse and filter every API in the catalog.
get_api_artifactsOne API's artifacts, grouped by type.
get_openapiThe primary OpenAPI for this API.
find_similar_apisAPIs that look like this one.
apis_io_searchSTART HERE — APIs, providers and tags for one query, each with its total.
resolveTurn a domain, URL or GitHub org into the provider it belongs to.
find_cohortsEvery scored population of providers in the catalog.
All 92 tools →
Call it yourself
curl for this page
This API
curl "https://apis.io/api/v1/apis/artificialwatch:artificialwatch-leaderboards-api"
All apis
curl "https://apis.io/api/v1/apis?limit=25"
Calls without X-API-Key need no key. Calls with it need a paid plan — get a key.
Get an API key
Free tier, no form to fill in. Signing in shares your email address with us — we
store it to create your key and to recognise you if you sign in with another
provider. See our Privacy Policy and
Terms.
A second provider on the same verified email joins the account you already have.
openapi: 3.2.0
info:
title: ArtificialWatch open data Leaderboards API
version: 1.0.0
summary: AI model releases and shutdowns, first-party LLM API prices, published knowledge cutoffs, open-weights models, OpenCode's stealth models and leaderboard indexes as static JSON, CSV and iCalendar files.
description: 'Read-only, keyless, CORS-open files at stable URLs: the AI model release log (dated by the day each model went live on a public API), each live model''s list price on its maker''s own API, the composite Bench and Coding indexes across public leaderboards, first-seen versus announcement times, every dated model shutdown from the labs'' own deprecation pages, the knowledge cutoff each lab publishes for its live models, every open-weights model with its Hugging Face repo, licence, size and rank, and every anonymous model OpenCode has listed with the lab that later claimed it. Files refresh with the site (about every 4 hours). Licence CC BY 4.0: credit ArtificialWatch with a link to the page named in each file''s `url`. Each board''s raw scores belong to that board.'
license:
name: CC BY 4.0
url: https://creativecommons.org/licenses/by/4.0/
contact:
name: ArtificialWatch
url: https://artificialwatch.com/data
servers:
- url: https://artificialwatch.com
security: []
tags:
- name: Leaderboards
description: Composite indexes across public leaderboards
paths:
/data/bench-index.json:
get:
operationId: getBenchIndex
summary: Bench Index
description: 'AI models ranked across the public LLM leaderboards: each board''s raw score, the model''s percentile on that board, and the composite index (median of percentiles).'
tags:
- Leaderboards
responses:
'200':
description: The Bench Index
content:
application/json:
schema:
$ref: '#/components/schemas/Index'
'503':
description: The table is rebuilding; retry shortly
/data/bench-index.csv:
get:
operationId: getBenchIndexCsv
summary: Bench Index (CSV)
description: 'The Bench Index as CSV: rank, id, name, vendor, index, boards, then _raw and _percentile per board.'
tags:
- Leaderboards
responses:
'200':
description: CSV with a header row
content:
text/csv:
schema:
type: string
'503':
description: The table is rebuilding; retry shortly
/data/coding-index.json:
get:
operationId: getCodingIndex
summary: Coding Index
description: The same composite across the public coding leaderboards.
tags:
- Leaderboards
responses:
'200':
description: The Coding Index
content:
application/json:
schema:
$ref: '#/components/schemas/Index'
'503':
description: The table is rebuilding; retry shortly
/data/coding-index.csv:
get:
operationId: getCodingIndexCsv
summary: Coding Index (CSV)
description: The Coding Index as CSV, in the Bench Index layout.
tags:
- Leaderboards
responses:
'200':
description: CSV with a header row
content:
text/csv:
schema:
type: string
'503':
description: The table is rebuilding; retry shortly
/data/open-weights.json:
get:
operationId: getOpenWeights
summary: Open-weights AI models
description: 'Every model whose weights are public on Hugging Face: the repo, its licence (standard or the lab''s own), total parameters, whether the repo is gated, and its rank against every model on the Bench and Coding indexes.'
tags:
- Leaderboards
responses:
'200':
description: The open-weights list
content:
application/json:
schema:
$ref: '#/components/schemas/OpenWeights'
'503':
description: The table is rebuilding; retry shortly
/data/open-weights.csv:
get:
operationId: getOpenWeightsCsv
summary: Open-weights AI models (CSV)
description: 'The open-weights list as CSV: id, name, lab, live_since, repo, hf_url, license, license_kind, gated, params, params_b, open_rank, bench_rank, bench_index, coding_rank, coding_index, checked, page.'
tags:
- Leaderboards
responses:
'200':
description: CSV with a header row
content:
text/csv:
schema:
type: string
'503':
description: The table is rebuilding; retry shortly
components:
schemas:
OpenWeightsModel:
type: object
required:
- id
- name
- lab
- live_since
- repo
- hf_url
- license
- license_kind
- gated
- params
- params_b
- open_rank
- bench_rank
- bench_index
- coding_rank
- coding_index
- checked
- page
properties:
id:
type: string
name:
type: string
lab:
type: string
live_since:
type:
- string
- 'null'
format: date
repo:
type: string
description: The Hugging Face repo, owner/name
hf_url:
type: string
format: uri
license:
type: string
description: The repo's own license tag; a lab's own licence reads custom (its name)
license_kind:
type: string
enum:
- standard
- custom
gated:
type: boolean
description: Hugging Face asks you to accept the lab's terms (and for a manual gate, wait for approval) before downloading
params:
type: string
description: Total parameters as Hugging Face counts them, e.g. 763B or 2.8T
params_b:
type:
- integer
- 'null'
description: The same size in billions
open_rank:
type:
- integer
- 'null'
description: Rank among the open-weights models
bench_rank:
type:
- integer
- 'null'
description: Rank against every model on the Bench Index, open and closed
bench_index:
type:
- number
- 'null'
coding_rank:
type:
- integer
- 'null'
description: Rank against every model on the Coding Index
coding_index:
type:
- number
- 'null'
checked:
type: string
format: date
page:
type: string
format: uri
Index:
type: object
required:
- name
- url
- license
- sources
- models
properties:
name:
type: string
url:
type: string
format: uri
license:
type: string
format: uri
attribution:
type: string
method:
type: string
description:
type: string
generatedAt:
type:
- string
- 'null'
sources:
type: array
items:
type: object
required:
- id
- name
- url
properties:
id:
type: string
name:
type: string
url:
type: string
format: uri
scale:
type: string
asOf:
type:
- string
- 'null'
models:
type: array
items:
type: object
required:
- rank
- id
- name
- vendor
- index
- boards
- scores
- page
properties:
rank:
type: integer
id:
type: string
name:
type: string
vendor:
type: string
index:
type: number
boards:
type: integer
scores:
type: object
description: Keyed by source id; null where the board does not rate the model
additionalProperties:
type:
- object
- 'null'
properties:
raw:
type: number
percentile:
type: number
page:
type: string
format: uri
OpenWeights:
type: object
required:
- name
- url
- license
- generated
- count
- models
properties:
name:
type: string
url:
type: string
format: uri
license:
type: string
format: uri
attribution:
type: string
method:
type: string
generated:
type: string
format: date
bench_generated:
type:
- string
- 'null'
format: date
ranked_of:
type:
- integer
- 'null'
description: Models on the Bench Index
coding_ranked_of:
type:
- integer
- 'null'
description: Models on the Coding Index
count:
type: integer
standard_license:
type: integer
description: Models under a standard licence (MIT, Apache 2.0)
models:
type: array
items:
$ref: '#/components/schemas/OpenWeightsModel'
externalDocs:
description: The open data page
url: https://artificialwatch.com/data