Your niches, briefed daily.

The starting point for your configured workspace — canonical, caller-owned, and legacy topic_id values plus coverage stats for each.

copycurl -s "https://api.metrix.ac/api/v1/niches" \
  -H "Authorization: Bearer $TM_API_KEY"

copy{
  "niches": [
    { "topic_id": "food_cooking", // ← pass this as {topic_id} below
      "name": "Food & Cooking", "category": "food",
      "seed_count": 123, "hashtag_count": 115,
      "hashtag_seeds": 115, "creator_seeds": 8,
      "video_count": 3587, "coverage": "good", "status": "ready" }
  ],
  "total": 3
}

seed_count is the total across every seed type. hashtag_count is retained as a hashtag-specific field and equals hashtag_seeds; it is not the total.

One request, one screen: every niche you can see, its current 24h velocity, which way it's trending, and its single top mover — sorted by ?sort= (default velocity_24h; also velocity_delta, opportunity_count).

copycurl -s "https://api.metrix.ac/api/v1/niches/compare" \
  -H "Authorization: Bearer $TM_API_KEY"

copy{
  "total": 3,
  "summaries": [
    { "topic_id": "food_cooking", "velocity_24h": 128541.43, "velocity_delta": 0.0,
      "trend_direction": "flat", // ← up / down / flat since the last compute
      "top_entity_id": "relationships", "top_entity_type": "hashtag", "opportunity_count": 0 }
  ]
}

Pass ?topic_ids= (comma-separated) to compare a subset of your registry-visible topics.

copycurl -s "https://api.metrix.ac/api/v1/niches/food_cooking/seeds" \
  -H "Authorization: Bearer $TM_API_KEY"

Unlike the directory's platform-validated term view, this response preserves all four configured seed types: hashtag, creator, sound, and comment_author.

copycurl -s -X POST "https://api.metrix.ac/api/v1/niches/food_cooking/affinity/recompute?entity_type=creator&min_evidence=1&limit=50" \
  -H "Authorization: Bearer $TM_API_KEY"

copycurl -s "https://api.metrix.ac/api/v1/niches/food_cooking/affinity?entity_type=creator&min_affinity=0.5&limit=50" \
  -H "Authorization: Bearer $TM_API_KEY"

Same velocity leaderboard as the trending page, pre-scoped to one niche. entity_type is one of hashtag (default), sound, creator, or ig_hashtag (Instagram — see the note above); window defaults to 24h.

copycurl -s "https://api.metrix.ac/api/v1/niches/food_cooking/trending?limit=3" \
  -H "Authorization: Bearer $TM_API_KEY"

copy# same niche, Instagram hashtag trends instead of TikTok
curl -s "https://api.metrix.ac/api/v1/niches/food_cooking/trending?entity_type=ig_hashtag&limit=3" \
  -H "Authorization: Bearer $TM_API_KEY"

copy{
  "entity_type": "hashtag",
  "window": "24h",
  "results": [
    { "entity_id": "relationships", "velocity": 128541.43, // ← the niche's #1 climber
      "acceleration": -12262.94, "size_tier": "xl", "rank_velocity": 10 },
    { "entity_id": "icecream", "velocity": 53176.04, "size_tier": "medium", "consecutive_positive_windows": 2 }
  ]
}

Filters to entities with positive acceleration that have climbed for two consecutive windows in a row — this is the "still moving" list rather than the raw leaderboard above. It spans every size tier by default; pass ?tier=nano for a nano-only early-signal view.

copycurl -s "https://api.metrix.ac/api/v1/niches/fitness_health/rising?entity_type=hashtag&limit=5" \
  -H "Authorization: Bearer $TM_API_KEY"

copy{
  "entity_type": "hashtag",
  "window": "24h",
  "results": [
    { "entity_id": "dance", "velocity": 362601.65, "acceleration": 31530.58, // ← still climbing, not just big
      "size_tier": "large", "consecutive_positive_windows": 4 }
  ]
}

Live-captured against fitness_health — food_cooking and brainrot_memes both came back empty for this window on every entity_type we tried. That's a real, expected result: this endpoint only surfaces entities with two straight windows of positive acceleration, and most windows most niches have none. Default entity_type is sound; we passed hashtag above to get a result to show.

Not a leaderboard sorted by a metric — a list of entities whose classified size_tier itself jumped by min_tier_jump or more steps (default 2) within the window. This is the closest thing to "went viral overnight" the API can point at directly.

copycurl -s "https://api.metrix.ac/api/v1/niches/brainrot_memes/breakouts?limit=5" \
  -H "Authorization: Bearer $TM_API_KEY"

copy{
  "topic_id": "brainrot_memes",
  "min_tier_jump": 2,
  "total": 1,
  "breakouts": [
    { "entity_id": "dancecompetition", "entity_type": "hashtag",
      "old_tier": "nano", "new_tier": "medium", // ← nano → medium in one window
      "tier_jump": 4 }
  ]
}

Live-captured against brainrot_memes — food_cooking and fitness_health both came back with total: 0 for this window. A 2+ tier jump is genuinely rare; an empty result here is the common case, not a broken call.

Looks across every other niche you can see, finds entities that are trending there, and checks whether this niche has touched them at all. What comes back is a shortlist of formats or hashtags proven to work elsewhere that no one in this niche has tried.

copycurl -s "https://api.metrix.ac/api/v1/niches/food_cooking/content-gaps?limit=5" \
  -H "Authorization: Bearer $TM_API_KEY"

copy{
  "topic_id": "food_cooking",
  "total": 3,
  "results": [
    { "entity_id": "comedy", "entity_type": "hashtag", "max_velocity": 3389019.52, // ← untouched here, huge elsewhere
      "source_niches": ["brainrot_memes", "fitness_health", "gaming", "sports"] }
  ]
}

source_niches can include niches outside your own subscription (they're the source of the trend, not something you're being shown data for) — trimmed above from 8 entries to 4.

affinity — how strongly a creator's actual posts overlap with this niche's signals, not just whether they've ever used one relevant hashtag. Higher affinity_score means more of their content genuinely belongs here.

copycurl -s "https://api.metrix.ac/api/v1/niches/food_cooking/creators?limit=3" \
  -H "Authorization: Bearer $TM_API_KEY"

copy{
  "topic_id": "food_cooking",
  "total": 3,
  "results": [
    { "creator_id": "ambreen_08", "affinity_score": 0.9999, // ← 0–1, how well they fit this niche
      "confidence": 0.0174, "velocity": null, "size_tier": null }
  ]
}

confidence is separate from affinity_score — a low confidence (like 0.0174 above) means the score is based on a thin sample of that creator's posts, so treat a high-affinity/low-confidence creator as a lead to verify, not a settled fact. velocity/size_tier come back null when we haven't computed a velocity row for that creator yet.

Pairs of hashtags that keep showing up together on this niche's top-performing videos, ranked by the average view count of videos using both. min_video_count (default 3) filters out one-off coincidences.

copycurl -s "https://api.metrix.ac/api/v1/niches/food_cooking/hashtag-combos?limit=3" \
  -H "Authorization: Bearer $TM_API_KEY"

copy{
  "topic_id": "food_cooking",
  "total": 3,
  "results": [
    { "hashtag_a": "fyp", "hashtag_b": "unitedstates", "video_count": 17,
      "avg_views": 188600000.0, // ← average views when both tags are used together
      "max_views": 372600000 }
  ]
}

Every (day, hour) slot this niche has posted in, with how many videos and their average views — sorted by average views, so the first row is this niche's best-performing posting slot.

copycurl -s "https://api.metrix.ac/api/v1/niches/food_cooking/timing" \
  -H "Authorization: Bearer $TM_API_KEY"

copy{
  "topic_id": "food_cooking",
  "total": 148,
  "slots": [
    { "day_of_week": 1, "hour_of_day": 20, "video_count": 24, "avg_views": 36344860.5 }, // ← this niche's best slot
    { "day_of_week": 5, "hour_of_day": 23, "video_count": 35, "avg_views": 17705726.8 }
  ]
}

day_of_week follows Postgres' EXTRACT(DOW) convention: 0 = Sunday … 6 = Saturday. hour_of_day is UTC, not the creator's local time.