Search Huberman Lab Health Topics

Purpose

Search hubermanlab.com for a health topic (e.g. "sleep", "sleep hygiene", "cold exposure", "dopamine") and return the matching episodes, newsletters, topic pages, and timestamped clips. This is a read-only lookup. The site's search is powered by a public, read-only Meilisearch index whose host/key/index are exposed in the page's client-side JavaScript, so the optimal path is a direct API call — no browser automation required. A browser fallback is documented for environments that can't reach the Meilisearch host directly.

When to Use

"Find Huberman Lab episodes about sleep / sleep hygiene."
"What has Andrew Huberman covered on cold exposure / caffeine / dopamine?"
"Search hubermanlab.com for {topic} and list the top results with links."
"Get the topic/category breakdown of Huberman Lab content matching {term}."
Building a content index or pulling episode metadata (titles, slugs, post dates, categories, timestamps) for a given health subject.

Workflow

The recommended method is a direct HTTP POST to the Meilisearch index that the site itself queries. The search box at https://www.hubermanlab.com/search-results is a thin client-side wrapper around this index — there is nothing the UI returns that the API doesn't.

Step 1 — Issue the search request

POST https://ms-3c8f24792474-17893.sfo.meilisearch.io/indexes/hubermanlab/search
Authorization: Bearer 041e82b255005543758795a8d50c273d19024ae10fddf69616ef1c8dc596711a
Content-Type: application/json

{
  "q": "sleep",
  "limit": 20,
  "offset": 0,
  "attributesToHighlight": ["name", "description"],
  "attributesToCrop": ["description"],
  "cropLength": 200
}

A GET form also works: GET /indexes/hubermanlab/search?q=sleep&limit=20 with the same Authorization: Bearer header.

Step 2 — (optional) Filter by category

Use Meilisearch facet filters to narrow by content type:

{ "q": "sleep", "limit": 20, "filter": "category = \"Solo Episode\"" }

Valid category values (with current counts from facet distribution): Solo Episode (101), Guest Episode (172), Essentials (72), AMA (19), Newsletter (19), Guest Series (16), Sponsor (16), Topic (29), Journal Club (2), Timestamp (2260). To get live facet counts, send { "q": "", "facets": ["category"], "limit": 0 } and read facetDistribution.category.

Step 3 — Parse results

Each element of hits[] contains: name, slug, link (relative path, prefix with https://www.hubermanlab.com), category, primaryTopic[], topics[], subtopics[], postDate, thumbnail, description, keywords, episodeTimestamps[], and _formatted (the same fields with <em>…</em> highlight markup). estimatedTotalHits is the total match count. Build the user-facing list from name + https://www.hubermanlab.com{link} + category.

Browser fallback

If the Meilisearch host is unreachable (network egress restrictions), drive the search page directly. The query parameter auto-runs the search on load:

Open https://www.hubermanlab.com/search-results?q=sleep (use a remote/stealth browser session — see Gotchas).
wait timeout 2500 — results are rendered client-side; the initial HTML is an empty shell (same byte length for every query).
snapshot and read the result cards (title, category badge, topic tags, link). Pagination shows ~20 hits/page.
To change the query without reloading, type into #search-input (debounced 500ms, minimum 3 chars) — but navigating to ?q={term} is simpler and deterministic.

Site-Specific Gotchas

The Meilisearch credentials are a public, read-only search key embedded in the page's own client JS (MEILISEARCH_HOST / MEILISEARCH_API_KEY / INDEX_NAME = "hubermanlab"). It only permits search, not writes. If the key rotates, re-extract it from the HTML of https://www.hubermanlab.com/search-results (search the page source for MEILISEARCH_HOST).
hubermanlab.com 301-redirects to www.hubermanlab.com (Cloudflare + Webflow). Always use the www. host. /search, /search?query=…, and /search?q=… all 301 to /search-results. The live query param on the results page is q (e.g. /search-results?q=sleep).
Search results are 100% client-side rendered. Fetching the page HTML returns the same 92 KB shell regardless of query — you will see zero results in raw HTML. You must either call the API (recommended) or render JS in a real browser.
The site's own UI uses Meilisearch hybrid search (hybrid: { embedder: "hubermanlab" }) for queries ≥ 3 chars. In testing the embedder returned semanticHitCount: 0 (no real vector matches), so for valid terms it behaves like keyword search. Avoid hybrid for clean no-result detection: a gibberish query with hybrid returns ~1000 nearest-neighbor junk hits, whereas plain keyword search correctly returns estimatedTotalHits: 0. Use plain keyword search unless you specifically want fuzzy/semantic recall.
The index mixes content types. Besides episodes/newsletters/topics, it contains 2260 Timestamp entries (individual clip markers). If you only want full episodes/articles, filter to the relevant categories (Solo Episode, Guest Episode, Essentials, etc.) and exclude Timestamp.
link is relative. Episode links look like /episode/{slug}; topic links like /topic/{slug}. Prefix with https://www.hubermanlab.com.
Cloudflare/anti-bot: the pre-run probe flagged Cloudflare and suggested residential proxies. The direct Meilisearch API call is on a separate host (*.meilisearch.io) with no Cloudflare and worked cleanly. For the browser fallback, a Browserbase remote session reached the page successfully even without proxies (built-in stealth was enough); residential proxies (--proxies) are a safe upgrade if you hit a challenge.
There is also an "Ask Huberman" AI feature linking to https://ai.hubermanlab.com/start?content={query} — not part of search; ignore unless explicitly asked.

Expected Output

Successful search (recommended API path):

{
  "success": true,
  "query": "sleep",
  "total_results": 249,
  "results": [
    {
      "name": "Sleep Toolkit: Tools for Optimizing Sleep & Sleep-Wake Timing",
      "category": "Solo Episode",
      "primary_topic": "Sleep Hygiene",
      "topics": ["Sleep Hygiene", "Light Exposure and Circadian Rhythm"],
      "post_date": "2022-08-08T08:00:00.000Z",
      "link": "https://www.hubermanlab.com/episode/sleep-toolkit-tools-for-optimizing-sleep-and-sleep-wake-timing",
      "description": "In this Huberman Lab episode, discover science-backed tools for optimizing sleep and sleep-wake timing..."
    }
  ]
}

Category-filtered search (filter: category = "Solo Episode"):

{ "success": true, "query": "sleep", "filter": "category = \"Solo Episode\"", "total_results": 37, "results": [ /* … */ ] }

No results (plain keyword search, no hybrid):

{ "success": true, "query": "zxqwplmkjhgf", "total_results": 0, "results": [] }

Facet distribution (q: "", facets: ["category"], limit: 0):

{
  "success": true,
  "facetDistribution": {
    "category": { "Solo Episode": 101, "Guest Episode": 172, "Essentials": 72, "AMA": 19, "Newsletter": 19, "Topic": 29, "Timestamp": 2260, "Guest Series": 16, "Sponsor": 16, "Journal Club": 2 }
  }
}

search-health-topics