Skip to content

Toolradar Research

Open source software statistics 2026

370 published Toolradar tools put open source in the name, the slug, or the tagline. 340 of them (92%) have a free path. Only 4 tools in the whole catalog carry the open_source pricing label.

Louis Corneloup
Louis Corneloup

Founder, Toolradar & Dupple

Published September 22, 2026
3 min read
Next update Dec 22, 2026
As featured inTechCrunchForbesBloombergBusiness InsiderThe Verge

Key findings

What the data shows.

  1. 01

    370 published tools have "open source" or "open-source" in the name, the slug, or the tagline. (Source: Toolradar directory, September 22, 2026.)

  2. 02

    Of those 370: 165 free, 173 freemium, 2 labeled opensource, 28 paid, 1 pay-per-use, 1 usage-based. 340 (92%) have a free path. 28 (8%) are labeled paid anyway. (Source: same query.)

  3. 03

    The pricing label opensource exists on 4 published tools in the whole catalog of 10,667. Two of those four are inside the 370 name-matches. The label is not the market. (Source: Toolradar directory, September 22, 2026.)

  4. 04

    This is a string match, the same kind of cut as MCP servers statistics. It is not a license audit. A tool can be GPL and never say so in the tagline. A tool can say open source and still charge.

  5. 05

    The numbers are not additive with MCP, AI Agents, or Developer Tools. A tool can sit in more than one cut.

About the research

How we built this report.

Data source

Toolradar tool database. Editorial review with weekly pricing verification.

Coverage period

2026. Snapshot taken September 22, 2026. Refresh due Dec 22, 2026.

Methodology

Public scoring rubric. See how we rate for the full criteria.

License

Creative Commons BY 4.0. Quote, link, and reuse with attribution.

How we counted

  1. Keep published tools only. The catalog that morning was 10,667.
  2. Keep a tool if the name, the slug, or the tagline matches "open source" or "open-source", case-insensitive.
  3. One row per tool. Result: 370.
  4. Split on the listing's pricing label.

We did not open GitHub licenses. We did not read COPYING files. PostgreSQL, MySQL, Bitwarden, and OBS Studio are in the matched set because the words are on the listing. A famous open-source project whose tagline never says the words is out. If a later pass adds a license field, this page should switch to that field and retire the string match.

Why the label and the words disagree

The pricing column has a value open_source. It is used 4 times. The words "open source" show up 370 times. Treating the label as the size of open source on Toolradar would publish 4 and hide the rest.

The 28 paid name-matches are the other direction. Saying the words does not mean the listing is free. A hosted open-source product can sell a cloud plan and keep the label paid. This page does not reclassify those 28. It counts them.

Free path here means the labels free, freemium, and open_source: 165 + 173 + 2 = 340. Pay-per-use and usage-based stay out of that sum. The same rule is used in free-tier share by category, where the catalog-wide free path is 5,834 of 10,667 (55%) on this same morning. Name-matched open source is far more often free-path than the average tool. It looks like the MCP cut (82% free path on 346 name-matches), and it is still a different set. Do not add 370 and 346.

What this page refuses

  • It will not say 370 tools are OSI-approved, GPL, or self-hostable.
  • It will not say the 4 open_source labels are the open-source market.
  • It will not turn the 28 paid rows into a price index. Dollar moves from the unreviewed pricing queue are in software pricing changes we will not print yet.

Buyer pages next to this cut: best open source developer tools, best open source databases.

Cite this report

Use the data, credit the source.

Released under Creative Commons BY 4.0. You may quote, link, and reuse the data with attribution.

Toolradar Research (2026). Open source software statistics 2026. Toolradar. https://toolradar.com/reports/open-source-software-statistics