Open Source Stats

verified statistics · every figure sourced · comparison tables, not single numbers
Home / Rankings

Target-query rankings

run 2026-10-24T11:27:38.412Z · 5 run(s) recorded · 0 of 13 checks answerable

What this measures

Whether a search backend returns this site in its first 20 results for a question the site is built to answer. Bing matters because it feeds ChatGPT and Copilot; Brave because it is Claude's search backend; Google because of AI Overviews.

— means the backend answered and this site was not in the top 20. blocked means the backend refused the request; decoy_results means it answered with results that do not address the query; no_results_parsed means the page yielded nothing parseable. None of those three is evidence about our position, and none is ever counted as absent. · means no data for that pair yet.

Measured by scraping, which mostly does not work. Bing answers automated clients with a decoy - a measured case returned ten "what is my IP" sites, and its RSS endpoint returned dictionary entries for the word "many", for a question about GitHub stars. Google walls headless clients behind consent. DuckDuckGo redirects to its homepage. Brave answers honestly until it rate limits. Real Bing and Google position data has to come from Webmaster Tools and Search Console, which need a verified domain; that data is imported here once it exists rather than guessed at.

This run

Ranked anywhere
0 of 0 answerable
Branded query ranked
0
Blocked / errored
13 of 13
A branded query is the cheapest possible win: someone types our name and the backend finds us. If that returns nothing, the site is not indexed, and no amount of content or comparison will change it until it is.

Queries × backends

IDQueryBucket BingBraveGoogle
Q01AI coding tools comparison stars forks
cross-tool table; no single repo page answers it
winnable···
Q02opencode vs claude code github stars
pairwise comparison across two owners
winnable···
Q03open source AI coding assistant comparison 2026
category roundup
winnable···
Q04fork to star ratio AI coding tools
derived metric, exists nowhere else
winnable···
Q05open issues count AI coding CLI comparison
maintenance signal, needs the join
winnable···
Q06AI coding agent popularity ranking github
ranking requires many repos at one timestamp
winnable···
Q07fastest growing AI coding tool
needs a growth series; we have 3 readings over 2 days
needs-history···
Q08which AI coding tool will overtake claude code
needs trend to extrapolate
needs-history···
Q09AI coding tool star growth rate 2026
needs history (IT-0215 backfill unblocks this)
needs-history···
Q10most popular AI coding tool trend october 2026
trend word requires a series
needs-history···
Q11how many github stars does opencode have
github.com answers in one hop; do not chase
never···
Q12Dataindex AI coding index
E1 term; measures whether our own name is retrievable
branded···
Q13dataindex github stars data
brand+category; the query a returning user would type
branded···
Buckets: winnable — this site can answer it and no single page else does; needs-history — requires a growth series we do not have yet; never — github.com answers it in one hop, so do not chase it; branded — our own name, the cheapest possible test.

Decision line

If no unbranded query reaches top 20 on any backend by D+45, expected fetches are ~0 and continuing the wait is not worth it.