Pisama Data Report · June 3, 2026

Most AI-platform adoption numbers are inflated. We counted the real ones.

Platform footprints are usually tallied by keyword mentions, which overstate genuine built-with adoption by anywhere from a few times to five hundred. We fingerprinted 505,406public repositories for the real markers, a lovable-tagger dependency, a .replit file, a langgraph import, and counted what is actually built with each AI app platform. Because those confirmed counts are conservative floors that miss private and deleted apps, we lead with the relative picture: each platform's share of new built-with apps, and its momentum against GitHub's own growth, month by month.

505,406
Repositories analyzed
83,048
Confirmed built-with
10
Platforms tracked
71,486
Distinct developers
Exhibit 1Hype versus realityKeyword mentions vs confirmed built-with
Per platform: keyword ecosystem (muted) against signature-confirmed built-with (amber)
Streamlit47,438 keyword · 36,105 built-with1.3×
LangGraph39,068 keyword · 21,680 built-with1.8×
Replit26,624 keyword · 16,290 built-with1.6×
n8n19,144 keyword · 3,950 built-with4.8×
Lovable15,124 keyword · 2,571 built-with5.9×
Gradio1,657 keyword · 963 built-with1.7×
Dify3,501 keyword · 896 built-with3.9×
Bolt711 keyword · 359 built-with2.0×
v0464 keyword · 209 built-with2.2×
Glide2,338 keyword · 25 built-with94×
openclaw25,940 keyword · 51 built-with509×

The clearest case of why keyword counts mislead. The name “openclaw” collides with a well-known open-source video game, so 25,940 repositories matched the term in March 2026 while only 51 carried a real build signature. Pisama covers OpenClaw as a framework, and the honest number is 51.

Inflation tracks name distinctiveness. Platforms with a unique fingerprint (Streamlit, the .replit file) sit near their real count; platforms with common names or noisy keywords run far ahead of what is actually built with them.

Exhibit 2The shifting mixShare of monthly built-with · 2025-06 to 2026-05
Each platform's share of all confirmed built-with apps, by month
0%30%61%
20252026
StreamlitLangGraphReplitn8nLovableGradioOther
Where things stand, in relative terms
Streamlit
27.8% now/60.7% a year ago/1.0× since
LangGraph
33.3% now/21.9% a year ago/3.5× since
Replit
31.6% now/4.3% a year ago/17× since
n8n
3.4% now/4.9% a year ago/1.6× since
Lovable
1.8% now/2.4% a year ago/1.8× since
Gradio
0.8% now/2.3% a year ago/0.8× since
Dify
0.9% now/1.4% a year ago/1.5× since
Bolt
0.2% now/1.8% a year ago/0.2× since
v0
0.2% now/0.3% a year ago/new
Glide
0.0% now/0.0% a year ago/new

Shares are robust where absolute counts are not. Streamlit still leads but its share is falling as Replit and LangGraph climb. “Now” is the latest complete month; “a year ago” is the first. The multiple is each platform measured against its own start.

Exhibit 3Momentum, against GitHub itselfIndexed to June 2025 = 100
Each platform against its own start, with GitHub's overall repo creation as the dashed baseline
08371,673
20252026
StreamlitLangGraphReplitn8nLovableGradioGitHub overall
Growth once GitHub's own rise is divided out
Replit9.2× GitHub's pace
LangGraph1.9× GitHub's pace
Lovable1.0× GitHub's pace
n8n0.9× GitHub's pace
Streamlit0.6× GitHub's pace
Gradio0.5× GitHub's pace

Measuring each platform against itself sidesteps the undercounting, but GitHub itself nearly doubled over the year, so the dashed baseline is the real bar to clear. Replit cleared it about nine times over and LangGraph nearly twice; n8n and Lovable merely kept pace with GitHub, while Streamlit and Gradio grew slower than the platform at large. Across every tracked platform, confirmed built-with still grew about 2.3 times against GitHub's 1.8, a slim but real gain in share of everything created. Platforms with too small a starting base to index are omitted.

FindingsWhat stands out
17×
Replit confirmed built-with apps grew seventeenfold in twelve months, the steepest momentum in the set.
about nine times GitHub's own growth over the same months
+81%
GitHub's own repo creation nearly doubled over the year. Adjusted for that baseline, only Replit and LangGraph genuinely outran the platform; Streamlit and Gradio grew slower than GitHub itself.
4.9M new repos in Jun 2025 to 8.8M in May 2026, same dated search
54%
of discovered Lovable repositories were deleted or made private within weeks, one reason absolute counts mislead.
794 of 1,681 May candidates still reachable
1.8%
Lovable share of new built-with apps in the latest month, well below its profile, because most of its apps never reach public GitHub.
a conservative floor, private and hosted apps are missing
ContextThe broader ecosystem505,406 repositories
Where the apps live, by source platform
GitHub455,259
Dev.to34,459
npm10,384
Lobsters1,627
Hacker News801
GitLab690
Docker Hub569
crates.io549
Packagist548
Rubygems194
Top languages, where detected
Python155,825
TypeScript81,551
JavaScript51,329
Jupyter Notebook24,823
HTML15,322
Shell7,625
Go7,089
Rust5,636
Java5,302
C#3,521

Raw discovery counts, shown for scale and context rather than ranking. The per-platform adoption story is in the relative exhibits above.

MethodHow we counted

Discovery. De-floored, date-windowed GitHub search across about 90 platform queries, 505,406 public repositories captured with their creation dates.

Built-with.Each candidate repository's files were fetched from the raw content CDN and confirmed only on a real fingerprint: a lovable-tagger dependency, a .replit file, a .bolt directory, a streamlit or langgraph import, or a platform project backlink.

Why relative. Confirmed counts are survivor lower bounds. Deleted or private repositories cannot be checked, and platforms whose apps are hosted off GitHub are undercounted. Shares and momentum are far more robust than absolute counts, so the charts lead with those.

The GitHub baseline. Total public repository creation on GitHub grew about 1.8 times over the year, from 4.9 million new repositories in June 2025 to 8.8 million in May 2026 by the same dated search. Momentum is read against that line, so outpacing the baseline is what counts, not simply rising with it.

The keyword trap. openclaw is the worked example: 25,940 keyword matches, 51 confirmed built-with, because the name collides with a video game.

Why PisamaThe same discipline, on running systems

This report exists because Pisama measures what software actually does, not what it claims. We build failure detection for multi-agent LLM systems, and we publish per-detector F1 scores and an open calibration dataset in a category that mostly publishes round numbers. The same discipline produced these counts. A fingerprint in the code, or it does not count.

See the per-detector F1 scores

Cite as: Pisama Built-With Report, June 3, 2026 (pisama.ai/stats). Press and data requests: [email protected].
Source data: ai-discovery-production · signature-confirmed built-with + de-capped catalogue. Aggregate, anonymized counts only. No individual app, author, or contact is identified.