Harpd Research · Diff engine
What changed on Harpd Rank: September 2026 vs August 2026
This page compares two closed monthly snapshots of the Harpd Rank board. Every number below is computed by differencing the two archive snapshots — board size, Rank Points, entrants, exits, rank movement and category composition. Nothing is estimated, and a metric without a comparable predecessor is shown as not comparable rather than as zero.
Rank Points are paid promotional placement, not an editorial quality score.
Quick answer
What changed on Harpd Rank between August 2026 and September 2026?
Between August 2026 and September 2026, the Harpd Rank board moved from 55 to 100 ranked products (+45), with total Rank Points +33. 99 products entered the board and 54 left it.
- Periods compared
- 2026-08 → 2026-09
- Ranked products
- 55 → 100 (+45)
- Total Rank Points
- 100 → 133 (+33)
- Products entered
- 99
- Products dropped
- 54
- License
- CC BY 4.0
Methodology & sources:Rank methodologyHistorical datasetResearch hub
Why it matters
Change is only measurable between two fixed points. Because each Harpd Rank month is a closed, immutable snapshot, this comparison stays valid after the live board moves on — you can cite what changed in a specific month without the numbers drifting underneath you.
Limitations
- Both periods must be closed UTC months with a materialized archive. An open month cannot appear here.
- Entrants and exits are set differences on product slug; a product that renames its slug is counted as one exit plus one entrant.
- Rank movement is only defined for products present in both snapshots. A product that entered has no previous rank and is not a mover.
- Rank Points are promotional placement purchased by makers, so a change in total Rank Points measures promotion, not market quality.
- Only products submitted to Harpd Rank appear. The board is not a census of the AI product market.
Data source: Harpd Rank closed monthly archive snapshotsLast updated: 2026-09
Board metrics
| Metric | 2026-08 | 2026-09 | Change |
|---|---|---|---|
| Ranked products | 55 | 100 | +45 |
| Total Rank Points | 100 | 133 | +33 (+33%) |
| Categories represented | 16 | 20 | +4 |
| Products entered | 99 | new | |
| Products dropped | 54 | left | |
Which products entered
| Product | Category | Rank (2026-09) | Rank Points |
|---|---|---|---|
| Armadai | developer | #2 | 0 |
| Blogged | seo | #4 | 0 |
| BuildAI.Space | other | #5 | 0 |
| Carillon | developer | #6 | 0 |
| Codeprepped | developer | #7 | 0 |
| CrawlRaven | other | #9 | 0 |
| Crew Tower | agents | #10 | 0 |
| Demand Discovery AI™ | developer | #11 | 0 |
| FeedFast | productivity | #14 | 0 |
| FounderStack | business | #15 | 0 |
| Hold My Lid | developer | #17 | 0 |
| HTML/CSS to Image API | ai-media | #18 | 0 |
| Huint | agents | #19 | 0 |
| Invofox | developer | #20 | 0 |
| LinkedGrow | agents | #21 | 0 |
| Lunaris | developer | #23 | 0 |
| Maileroo | developer | #24 | 0 |
| Metrician | developer | #25 | 0 |
| OutlierKit | other | #28 | 0 |
| Peppyhop | other | #29 | 0 |
Which products dropped
| Product | Category | Rank (2026-08) |
|---|---|---|
| Claude | agents | #2 |
| Perplexity | agents | #3 |
| Gemini | agents | #4 |
| Poe | agents | #5 |
| LangChain | agents | #6 |
| LlamaIndex | agents | #7 |
| CrewAI | agents | #8 |
| n8n | agents | #9 |
| Cursor | developer | #10 |
| GitHub Copilot | developer | #11 |
| Windsurf | developer | #12 |
| Replit | developer | #13 |
| Lovable | developer | #14 |
| Bolt | developer | #15 |
| Vercel | developer | #16 |
| Netlify | developer | #17 |
| Cloudflare | developer | #18 |
| Supabase | developer | #19 |
| Firebase | developer | #20 |
| Sentry | developer | #22 |
Which products moved
Movement is computed only for products present in both snapshots.
Biggest gains
No product gained rank between the two snapshots.
Biggest declines
No product lost rank between the two snapshots.
Which categories grew
| Category | 2026-08 products | 2026-09 products | Change | Share change |
|---|---|---|---|---|
| Other | 0 (0%) | 30 (30%) | +30 entered | +30 pp |
| Developer | 16 (29.1%) | 22 (22%) | +6 grew | -7.1 pp |
| Marketing | 1 (1.8%) | 5 (5%) | +4 grew | +3.2 pp |
| Education | 0 (0%) | 3 (3%) | +3 entered | +3 pp |
| Business | 1 (1.8%) | 3 (3%) | +2 grew | +1.2 pp |
| Audio | 0 (0%) | 2 (2%) | +2 entered | +2 pp |
| Writing | 0 (0%) | 2 (2%) | +2 entered | +2 pp |
| AI Media | 6 (10.9%) | 7 (7%) | +1 grew | -3.9 pp |
| Design | 1 (1.8%) | 2 (2%) | +1 grew | +0.2 pp |
| Hiring | 1 (1.8%) | 2 (2%) | +1 grew | +0.2 pp |
| Travel | 0 (0%) | 1 (1%) | +1 entered | +1 pp |
| News | 0 (0%) | 1 (1%) | +1 entered | +1 pp |
| Sales | 0 (0%) | 1 (1%) | +1 entered | +1 pp |
| Social | 0 (0%) | 1 (1%) | +1 entered | +1 pp |
| Productivity | 5 (9.1%) | 4 (4%) | -1 shrank | -5.1 pp |
| Leaderboards | 1 (1.8%) | 0 (0%) | -1 exited | -1.8 pp |
| Domains | 1 (1.8%) | 0 (0%) | -1 exited | -1.8 pp |
| Ecommerce | 1 (1.8%) | 0 (0%) | -1 exited | -1.8 pp |
| Agents | 9 (16.4%) | 7 (7%) | -2 shrank | -9.4 pp |
| Agencies | 2 (3.6%) | 0 (0%) | -2 exited | -3.6 pp |
| SEO | 6 (10.9%) | 3 (3%) | -3 shrank | -7.9 pp |
Gaining share: Other, Developer, Marketing, Education, Business. Losing share: Productivity, Leaderboards, Domains, Ecommerce, Agents.
Changes by research family
Each family is a segment view of the same board. Follow a family to its monthly report for the full breakdown.
| Family | 2026-08 | 2026-09 | Entered | Dropped |
|---|---|---|---|---|
| AI tools index | 37 | 47 | 46 | 36 |
| AI tools trends | 37 | 47 | 46 | 36 |
| Developer tools index | 16 | 22 | 22 | 16 |
| AI agent index | 9 | 7 | 7 | 9 |
| AI market index | 55 | 100 | 99 | 54 |
Method
The diff is computed by comparing two closed monthly archive snapshots of the Harpd Rank board (/data/rank-history.json). Each snapshot is materialized when its UTC month closes and never rewritten afterwards. Entrants and exits are set differences on product slug; movement is the signed rank delta for products present in both; category share is the category's product count over the board total for that period. Every figure on this page is derived from those two snapshots — see the Rank methodologyfor how positions are produced in the first place.
Cite this comparison
Both snapshots are immutable, so a reader who follows your citation sees the same two periods you compared. Cite the specific months — not "Harpd Rank" generally.
Cite this report
Data: https://harpd.com/data/rank-history.json
Method: https://harpd.com/rank/methodology/
Evidence: https://harpd.com/data/evidence.json
Updated: 2026-09-01
Licensed under CC BY 4.0. Attribution to Harpd is required when reused.
Machine-readable: /data/rank-history.json ·/data/citations.json ·how to cite Harpd · licence CC BY 4.0
Frequently asked questions
How often is this comparison updated?
Once per month, as soon as a new UTC month closes and its archive snapshot is materialized. Until then the page keeps showing the two most recent closed months.
Can a past comparison change?
No. Both snapshots are immutable closed-month archives, so a diff computed from them is fixed. This page always describes two specific periods, not the current live board.
Does a product leaving the board mean it was delisted?
Not necessarily. An exit means the product was present in the earlier closed snapshot and absent from the later one — the listing may have been removed, expired or renamed.
Why is a metric shown as "not comparable"?
Because its baseline is zero or absent. Reporting a percentage change against no data would be misleading, so it is reported as not comparable instead.