> ## Documentation Index
> Fetch the complete documentation index at: https://docs.cryptoquant.com/llms.txt
> Use this file to discover all available pages before exploring further.
# Analyst Track Record
> One analyst's accuracy record — scores and ranks, tier per asset class, call counts by source platform, recent bias, covered assets and where they publish. Public, no API key.
## What it returns
One analyst's record: `accuracy_score`, `accuracy_sample_size`, bull / bear / short / long scores
with their ranks, `overall_rank`, the tier per asset class, call counts split by source platform
(`calls.by_source`) with the definition of a call, recent bias, covered assets, the profile URL and
where they publish.
`{handle}` is the X handle, the platform handle or the `/analyst/{slug}` page slug.
## Access
Public — no API key, no session. Responses are CDN-cached for one hour and carry `score_notes`
explaining every number. Scores are comparative 0–100 values, not win probabilities — see the
[Methodology](https://consensus.cryptoquant.com/methodology).
## Notes
* A **call** is one directional (bullish or bearish) item about one tracked asset, over the whole
stored history: X posts, TradingView ideas, Seeking Alpha articles (title + summary), Substack
posts and sell-side analyst ratings. An article on several tickers is one call per ticker on each
asset page and counts once per article × ticker pair in its author's total.
* `calls.by_source` keys are the [`source_platform`](/sentiment/enums) values; `sources[]` uses
the `analyst_source` values (`twitter` = X).
* `tier` / `tiers` carry the badge key (`top1` · `top5` · `top10` · `tracked`), the pool rank and
size, and a `reason` when the tier is `null` — see [`tier_reason`](/sentiment/enums).
* No tracked analyst matching `{handle}` answers `404 ANALYST_NOT_FOUND`.
Building an agent tool? The same call backs the CryptoQuant MCP tool `get_analyst_profile` — see
[For AI Agents](/sentiment/for-agents).
## OpenAPI
````yaml openapi/sentiment.json GET /analysts/{handle}
openapi: 3.1.0
info:
title: Analyst Consensus API — CryptoQuant
version: 2.2.0
summary: >-
What curated analysts say about crypto, stocks, indices and commodities, as
numbers.
description: >-
What curated analysts on X, CryptoQuant, Seeking Alpha, TradingView and
Substack (plus licensed sell-side ratings on equities) say about 1712 assets
— crypto, US and KR equities, index ETFs, commodities — as numbers: the
daily Analyst Consensus Index (−100 … +100, one series per asset), a
per-asset consensus breakdown (bull / bear case, viewpoints grouped by
thesis, source posts), per-analyst stance series, and each analyst's
accuracy track record.
**Two access levels.** Public, no key (per-IP rate limit, CDN-cached): `GET
/assets` (the asset registry + every enum), `GET /consensus` without a key
(the LATEST index point for any asset with the bull / bear case), `GET
/analyst-views`, `GET /calls`, `GET /analysts/top`, `GET
/analysts/{handle}`. Every public payload carries `page_url` (link it),
`cite_as` and `data_by` ("Data by CryptoQuant Consensus" — show it with the
numbers). With an API key (`X-API-Key` header or `api_key` query parameter):
the index HISTORY (`GET /consensus` with `days`) and `GET
/consensus/breakdown` on any plan; the per-analyst series `GET /sentiment`
on the Premium and Enterprise plans only (`PLAN_REQUIRED` below). Keys are
provisioned with the CryptoQuant Premium plan or an Enterprise agreement;
there is no self-serve key endpoint on this host. **Raw source text** (the
original post title / text) is provided under an Integration agreement only,
per API key (`api_keys.raw_access`, never a plan): `GET
/consensus/breakdown` sends `sources[].statement` on a cleared key and a ≤
200-character `sources[].excerpt` + url on every other key — Premium,
Enterprise and admin-owned keys included; the key-free `/analyst-views` and
`/calls` excerpts follow the same rule. Integration use that needs the
original text: contact sales@cryptoquant.com (`x-raw-source-text`). Thesis,
narratives, index and labels are the same on every plan.
**Conventions.** Dates: `date` = UTC calendar day `YYYY-MM-DD`; `date-time`
= RFC 3339 / ISO 8601 with offset, UTC. Index and stance series are
end-of-day: the current UTC day is excluded until complete, so values never
change retroactively. No pagination anywhere: a series comes back whole
(`days` is capped by the plan, 36,500 = everything; the full BTC history is
≈ 1,950 rows); lists are capped by `limit` (1–50). Sorting is fixed per
endpoint and stated in its description. Filters combine with AND. Every
error is the `Error` object: branch on `code`, read `message`, follow
`docs_url`; `x-errors` lists every code with its recovery. Every operation
carries `x-plan` (who may call it).
**Agents.** Import this document as a ChatGPT GPT Action (Authentication:
None for the public operations) or any OpenAPI tool loader; the Markdown
version of these docs is `/llms.txt`; the same data is an MCP server at
`/mcp` (Streamable HTTP, no auth). Pass `via=chatgpt|claude|…` so the links
you show carry the right `utm_source`.
**Freshness.** Index and stance series: end-of-day UTC. The breakdown is
recomputed daily (English by 04:00 UTC, translations by 06:00 UTC); its
opinion counts refresh every 4 hours. Public payloads are CDN-cached 5
minutes (track record and assets: 1 hour).
Human docs: https://consensus.cryptoquant.com/docs/api · how the numbers are
made: https://consensus.cryptoquant.com/methodology · Markdown for agents:
https://consensus.cryptoquant.com/llms.txt
contact:
name: CryptoQuant — API access and Enterprise data
url: https://cryptoquant.com/get-in-touch
email: support@cryptoquant.com
termsOfService: https://consensus.cryptoquant.com/terms
license:
name: Proprietary — CryptoQuant Terms of Service
url: https://consensus.cryptoquant.com/terms
servers:
- url: https://consensus.cryptoquant.com/api/v1
description: Analyst Consensus by CryptoQuant
security:
- ApiKeyHeader: []
- ApiKeyQuery: []
tags:
- name: Registry
description: Public, no API key. The assets the API covers and every enum it uses.
- name: Consensus
description: >-
The Analyst Consensus Index: latest point public (no key); history and the
per-asset breakdown with an API key.
- name: Analyst views
description: >-
Public, no API key. What analysts are saying: narratives, per-analyst
views, the newest calls.
- name: Analyst track record
description: >-
Public, no API key. Who to trust: accuracy scores, ranks and badges per
analyst.
- name: Analyst stance
description: Premium / Enterprise API key. One analyst's daily stance on an asset.
externalDocs:
description: Sentiment Data documentation (the human rendering of this document)
url: https://docs.cryptoquant.com/sentiment/overview
paths:
/analysts/{handle}:
get:
tags:
- Analyst track record
summary: One analyst's track record (public, no API key)
description: >-
The analyst page's numbers as JSON: stored accuracy scores, category
ranks, overall rank, best top-10 placement, call counts split per source
and recent bias, covered assets, profile and source URLs. `{handle}` is
the X handle, the /analyst/`` page slug or a `platform:slug` id; a
leading @ is ignored and URL-encoding is decoded. Hidden, ineligible,
company, news-feed and sell-side rows are 404 `ANALYST_NOT_FOUND`.
Public; CDN-cached for one hour. Backs the MCP tool get_analyst_profile.
operationId: getAnalystTrackRecord
parameters:
- name: handle
in: path
required: true
schema:
type: string
example: caprioleio
description: X handle / page slug / `platform:slug` id (case-insensitive).
responses:
'200':
description: Track record
content:
application/json:
schema:
$ref: '#/components/schemas/AnalystTrackRecord'
example:
handle: caprioleio
name: Charles Edwards
slug: caprioleio
profile_url: https://consensus.cryptoquant.com/analyst/caprioleio
source_url: https://x.com/caprioleio
sources:
- twitter
role: Founder & CEO
company: Capriole Investments
follower_count: 0
is_verified: false
category: macro
activity:
last_activity_at: '2026-10-02T12:34:42+00:00'
active_last_3_months: true
track_record:
accuracy_score: 70.8
accuracy_sample_size: 1624
call_count: 2276
bull_score: 89.8
bear_score: 39.8
short_score: 63.9
long_score: 77
balanced_score: 39.8
overall_rank:
rank: 6
total: 109
category_ranks:
bull:
rank: 1
total: 111
bear:
rank: 51
total: 111
short:
rank: 20
total: 91
long:
rank: 29
total: 111
best_ranking:
category: bull
rank: 1
label: Bull Accuracy
priority: 2
scored: true
tier:
asset_class: crypto
tier: 1
badge: top1
label: Top 1%
eligible: true
reason: null
pool_rank: 7
pool_size: 106
percentile: 6.6038
score_version: v20r-btc-amp31-dir
computed_at: '2026-10-02T21:16:07.550877+00:00'
tiers:
crypto:
asset_class: crypto
tier: 1
badge: top1
label: Top 1%
eligible: true
reason: null
pool_rank: 7
pool_size: 106
percentile: 6.6038
score_version: v20r-btc-amp31-dir
computed_at: '2026-10-02T21:16:07.550877+00:00'
calls:
total: 1778
bullish: 1160
bearish: 618
neutral: 542
by_source:
x: 1778
cryptoquant: 0
news: 0
tradingview: 0
seekingalpha: 0
substack: 0
benzinga: 0
other: 0
definition: >-
A call is one directional (bullish or bearish) item about
one tracked asset, over the whole stored history: X posts,
TradingView ideas, Seeking Alpha articles (title + summary),
Substack posts and sell-side analyst ratings. An article on
several tickers is one call per ticker on each asset page
and counts once per article × ticker pair in its author's
total.
recent_bullish_bias_pct: 70
recent_bias: bullish
covered_assets:
- symbol: BTC
asset_class: crypto
last_post_at: '2026-09-23T04:02:49+00:00'
- symbol: SPY
asset_class: indices
last_post_at: '2026-08-05T06:40:15+00:00'
citation_count: null
score_notes:
accuracy_score: >-
Overall accuracy, 0–100: an Empirical-Bayes-adjusted hit
rate of the analyst's directional calls against subsequent
price moves, shrunk toward the pool mean when the sample is
small. Comparative — rank analysts on the same asset by it;
it is NOT a win probability.
accuracy_sample_size: >-
Number of evaluated calls behind accuracy_score. Treat
scores with a sample below ~10 as weak evidence.
methodology_url: https://consensus.cryptoquant.com/methodology
docs_url: >-
https://consensus.cryptoquant.com/docs/api#analyst-track-record
as_of: '2026-10-06T07:48:28.011Z'
'404':
description: >-
`ANALYST_NOT_FOUND` — no such tracked analyst (`message` points to
/analysts/top for valid handles)
content:
application/json:
schema:
$ref: '#/components/schemas/Error'
example:
error: Analyst not found
code: ANALYST_NOT_FOUND
message: >-
No tracked analyst "nobody". Valid handles are listed by GET
https://consensus.cryptoquant.com/api/v1/analysts/top?asset=.
docs_url: https://consensus.cryptoquant.com/docs/api#errors
security: []
x-codeSamples:
- lang: Shell
source: >-
curl -X GET
"https://consensus.cryptoquant.com/api/v1/analysts/caprioleio"
- lang: JavaScript
source: >-
fetch("https://consensus.cryptoquant.com/api/v1/analysts/caprioleio")
.then(response => response.json())
.then(data => console.log(data))
- lang: NodeJS
source: |-
require('axios')
.get("https://consensus.cryptoquant.com/api/v1/analysts/caprioleio")
.then(response => console.log(response))
- lang: Ruby
source: >-
require 'net/http'
uri =
URI("https://consensus.cryptoquant.com/api/v1/analysts/caprioleio")
res = Net::HTTP.get_response(uri)
puts res.body
- lang: Python
source: |-
import requests
url = "https://consensus.cryptoquant.com/api/v1/analysts/caprioleio"
print(requests.get(url).json())
components:
schemas:
AnalystTrackRecord:
type: object
required:
- handle
- slug
- profile_url
- sources
- activity
- track_record
- calls
- covered_assets
- as_of
properties:
handle:
type: string
description: X handle, or `:` for a platform-only analyst.
name:
type:
- string
- 'null'
profile_url:
type: string
format: uri
description: The analyst page on the site (cite this).
source_url:
type:
- string
- 'null'
format: uri
description: >-
Where the analyst publishes (X, CryptoQuant, TradingView, Seeking
Alpha, Substack).
sources:
type: array
items:
type: string
enum:
- twitter
- cryptoquant
- seekingalpha
- tradingview
- substack
- other
description: Stored provenance of the analyst row (`twitter` = X).
role:
type:
- string
- 'null'
company:
type:
- string
- 'null'
category:
type:
- string
- 'null'
description: >-
Analyst category label (e.g. `technical`, `macro`); free text, null
when unset.
slug:
type: string
description: Page slug (`/analyst/`).
follower_count:
type:
- integer
- 'null'
minimum: 0
is_verified:
type: boolean
activity:
type: object
required:
- last_activity_at
- active_last_3_months
properties:
last_activity_at:
type:
- string
- 'null'
format: date-time
active_last_3_months:
type:
- boolean
- 'null'
description: >-
null = no activity on record (new or seed-only analyst), not
"inactive".
track_record:
type: object
required:
- accuracy_score
- call_count
- overall_rank
- category_ranks
- scored
- tier
- tiers
properties:
accuracy_score:
type:
- number
- 'null'
minimum: 0
maximum: 100
description: >-
Overall accuracy, 0–100: Empirical-Bayes-adjusted hit rate of
directional calls vs. subsequent price, shrunk toward the pool
mean for small samples. Comparative within an asset; NOT a win
probability.
accuracy_sample_size:
type:
- integer
- 'null'
minimum: 0
description: >-
Evaluated calls behind accuracy_score; below ~10 is weak
evidence.
call_count:
type:
- integer
- 'null'
minimum: 0
description: >-
All-time directional (bullish or bearish) posts on tracked
assets.
bull_score:
type:
- number
- 'null'
minimum: 0
maximum: 100
description: Accuracy, 0–100, on calls made in bull market phases.
bear_score:
type:
- number
- 'null'
minimum: 0
maximum: 100
description: Accuracy, 0–100, on calls made in bear market phases.
short_score:
type:
- number
- 'null'
minimum: 0
maximum: 100
description: Accuracy, 0–100, of calls evaluated over the short horizon.
long_score:
type:
- number
- 'null'
minimum: 0
maximum: 100
description: Accuracy, 0–100, of calls evaluated over the long horizon.
balanced_score:
type:
- number
- 'null'
minimum: 0
maximum: 100
description: Mean of bull_score and bear_score when both exist.
overall_rank:
type:
- object
- 'null'
properties:
rank:
type: integer
minimum: 1
total:
type: integer
minimum: 1
description: Rank by accuracy_score among every analyst holding one.
category_ranks:
type: object
properties:
bull:
type:
- object
- 'null'
properties:
rank:
type: integer
minimum: 1
total:
type: integer
minimum: 1
bear:
type:
- object
- 'null'
properties:
rank:
type: integer
minimum: 1
total:
type: integer
minimum: 1
short:
type:
- object
- 'null'
properties:
rank:
type: integer
minimum: 1
total:
type: integer
minimum: 1
long:
type:
- object
- 'null'
properties:
rank:
type: integer
minimum: 1
total:
type: integer
minimum: 1
best_ranking:
type:
- object
- 'null'
properties:
category:
type: string
enum:
- bull
- bear
- short
- long
- overall
rank:
type: integer
minimum: 1
label:
type: string
priority:
type: integer
description: >-
The analyst's best top-10 placement across the categories, if
any.
scored:
type: boolean
tier:
description: >-
Headline badge: the best tier across the analyst's scored asset
classes (1 < 5 < 10, then class order). null = nothing stored
for the analyst (show "Tracked").
anyOf:
- type: object
allOf:
- $ref: '#/components/schemas/AccuracyTier'
- type: 'null'
tiers:
type: object
additionalProperties:
$ref: '#/components/schemas/AccuracyTier'
description: >-
Every asset class with a stored tier row, keyed by class. A
class absent here is unscored for this analyst.
calls:
type:
- object
- 'null'
required:
- total
- bullish
- bearish
- neutral
- by_source
- definition
- recent_bullish_bias_pct
- recent_bias
properties:
total:
type: integer
minimum: 0
bullish:
type: integer
minimum: 0
bearish:
type: integer
minimum: 0
neutral:
type: integer
minimum: 0
description: Neutral labelled posts — reported for context, never a call.
by_source:
type:
- object
- 'null'
description: >-
The analyst's directional calls per source over the whole stored
history (sums to `total`). `benzinga` is always 0 for a tracked
analyst (sell-side authors are not on the roster). null while
the breakdown function is not on the database.
required:
- x
- cryptoquant
- news
- tradingview
- seekingalpha
- substack
- benzinga
- other
properties:
x:
type: integer
minimum: 0
cryptoquant:
type: integer
minimum: 0
news:
type: integer
minimum: 0
tradingview:
type: integer
minimum: 0
seekingalpha:
type: integer
minimum: 0
substack:
type: integer
minimum: 0
benzinga:
type: integer
minimum: 0
other:
type: integer
minimum: 0
definition:
type: string
description: >-
What one call is: one directional (bullish or bearish) item
about one tracked asset, every source, whole history; an article
on several tickers counts once per article × ticker pair.
recent_bullish_bias_pct:
type:
- number
- 'null'
minimum: 0
maximum: 100
description: >-
Bullish share of the last 30 directional posts within three
months.
recent_bias:
type:
- string
- 'null'
enum:
- bullish
- bearish
- balanced
- null
covered_assets:
type: array
items:
type: object
required:
- symbol
- asset_class
- last_post_at
properties:
symbol:
type: string
asset_class:
type: string
enum:
- crypto
- equities
- indices
- commodities
last_post_at:
type:
- string
- 'null'
format: date-time
description: Most recent post first.
citation_count:
type:
- integer
- 'null'
minimum: 0
score_notes:
type: object
additionalProperties:
type: string
methodology_url:
type: string
format: uri
docs_url:
type: string
format: uri
as_of:
type: string
format: date-time
Error:
type: object
description: >-
Every non-2xx body. Branch on `code`; `error` is the legacy short reason
kept for pre-2026-10-06 clients.
required:
- error
- code
- message
- docs_url
properties:
error:
type: string
description: >-
Legacy short reason ("Invalid asset", "API key required", "Rate
limit exceeded" …), unchanged wording.
code:
type: string
enum:
- API_KEY_REQUIRED
- INVALID_API_KEY
- PLAN_REQUIRED
- RAW_ACCESS_REQUIRED
- DAILY_LIMIT_EXCEEDED
- RATE_LIMIT_EXCEEDED
- INVALID_PARAMETER
- UNKNOWN_ASSET
- ANALYST_NOT_FOUND
- ANALYST_NOT_ELIGIBLE
- NO_DATA
- SERVICE_UNAVAILABLE
- INTERNAL_ERROR
description: >-
Stable machine code. Status and recovery per code: see `x-errors` at
the document root.
message:
type: string
description: Human explanation, with the accepted values for a parameter error.
docs_url:
type: string
format: uri
description: >-
Where the error is documented (/docs/api#errors, or #plans for
PLAN_REQUIRED).
param:
type: string
description: The offending query parameter (INVALID_PARAMETER, UNKNOWN_ASSET).
required_plans:
type: array
items:
type: string
enum:
- premium
- enterprise
description: Plans that may call the endpoint (PLAN_REQUIRED).
plan:
type: string
description: The key's current plan label, upper-cased (PLAN_REQUIRED).
upgrade_url:
type: string
format: uri
description: Self-serve upgrade (PLAN_REQUIRED, DAILY_LIMIT_EXCEEDED).
contact_url:
type: string
format: uri
description: Enterprise / sales contact (PLAN_REQUIRED, DAILY_LIMIT_EXCEEDED).
limit:
type:
- integer
- 'null'
description: Daily quota (DAILY_LIMIT_EXCEEDED).
used:
type:
- integer
- 'null'
description: Requests used today (DAILY_LIMIT_EXCEEDED).
reset:
type: string
description: When the quota resets ("Daily at midnight UTC").
retry_after:
type: integer
minimum: 1
description: Seconds to wait (RATE_LIMIT_EXCEEDED; also the Retry-After header).
docs:
type: string
format: uri
deprecated: true
description: Legacy alias of docs_url on API_KEY_REQUIRED.
upgradeUrl:
type: string
format: uri
deprecated: true
description: Legacy alias of upgrade_url on PLAN_REQUIRED.
AccuracyTier:
type: object
description: >-
Accuracy percentile-tier badge for one asset class (Ki 2026-09-30). What
people see is the badge, stated against all analysts: tier 1 = "Top 1%"
(top 10% of the eligible tracked pool), 5 = "Top 5%" (top 50%), 10 =
"Top 10%" (the rest of the eligible pool); null = "Tracked" (no badge,
`reason` says why). Computed daily from the stored accuracy score; the
precise position (pool_rank / pool_size / percentile) is for detail
views and agents only.
required:
- asset_class
- tier
- badge
- label
- eligible
- reason
properties:
asset_class:
type: string
enum:
- crypto
- equities
- indices
- commodities
- macro
tier:
type:
- integer
- 'null'
enum:
- 1
- 5
- 10
- null
description: >-
Badge number: 1 = Top 1%, 5 = Top 5%, 10 = Top 10%; null = no badge
(Tracked).
badge:
type: string
enum:
- top1
- top5
- top10
- tracked
description: Stable key for the UI (i18n / icon).
label:
type: string
enum:
- Top 1%
- Top 5%
- Top 10%
- Tracked
description: English badge text.
eligible:
type: boolean
description: >-
True when the analyst is inside the class pool (scored, not
excluded, ≥ 30 directional calls). A thin pool (< 10) keeps
eligible=true with tier null and reason pool_too_small.
reason:
type:
- string
- 'null'
enum:
- excluded
- ineligible
- news_feed
- company
- unscored
- insufficient_calls
- pool_too_small
- null
description: >-
Why tier is null; null when a tier is set. unscored = no accuracy
score for this asset class.
pool_rank:
type:
- number
- 'null'
minimum: 1
description: >-
Midrank position from the top inside the eligible pool (1 = best;
tied analysts share one position, e.g. 2.5).
pool_size:
type:
- integer
- 'null'
minimum: 0
description: Eligible analysts in the class when computed.
percentile:
type:
- number
- 'null'
minimum: 0
maximum: 100
description: >-
100 × pool_rank / pool_size: position from the top in percent (0 =
best).
score_version:
type:
- string
- 'null'
description: >-
Score source the tier was computed from (`v20`, or an active
alternative such as `v20r-btc-amp31-dir`).
computed_at:
type:
- string
- 'null'
format: date-time
securitySchemes:
ApiKeyHeader:
type: apiKey
in: header
name: X-API-Key
description: >-
Provisioned with the CryptoQuant Premium plan or an Enterprise agreement
(no self-serve). Keys start with `unbias_live_`.
ApiKeyQuery:
type: apiKey
in: query
name: api_key
description: Same key as a query parameter (avoid in logged URLs).
````
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.