Current AI's Gap Map offers a structured dataset for navigating open source AI

The non-profit Current AI has released a map of 421 curated open source AI projects, backed by a public dataset of over 16,000 repos. We evaluate its utility for founders and investors.

The Answer Up Front

This is a powerful intelligence tool for founders, builders, and investors actively working in the open source AI ecosystem. It provides a structured, queryable dataset that goes far beyond typical "awesome list" repositories. Use it to find dependencies, scout for acquisition targets, or identify underserved market niches. You should skip it if you're looking for a simple, curated list of the "best" tools without wanting to perform your own analysis. The bottom line is that the Open Source AI Gap Map's primary value is its machine-readable, MIT-licensed dataset, not its web-based visualization. Its utility is directly proportional to your willingness to query the raw data.

Methodology

This review covers the Open Source AI Gap Map v0.1, as announced and observed in July 2026. The primary source is Simon Willison's analysis, which points to the official announcement from Current AI, the map itself, and the underlying public GitHub repository. This is a v0 review based on these initial public artifacts.

Our assessment focuses on the structure and potential utility of the dataset for its target audience of builders and investors. We explored the currentai-org/os-ai-map repository, examining the YAML file schemas and the provided CSV of GitHub repositories. We did not independently verify the categorization or completeness of the 421 curated projects or the 16,000+ uncategorized repositories. This review evaluates the Gap Map as a resource, not the individual tools it contains. Future updates will track the project's progress in categorizing its long tail of projects and implementing its proposed scoring system.

What It Does

A curated map of 421 projects

The project's public face is a web-based visualization of 421 hand-picked open source AI projects. According to the launch post, this initial set includes 266 software tools, 85 models, 50 datasets, and 20 hardware projects from 228 different organizations. These are organized across three layers of the AI stack: model components, product/UX, and infrastructure. This curated view serves as the high-level introduction to the ecosystem as defined by Current AI.

An open, queryable dataset

The most significant part of the project is the underlying data, published under an MIT license in a GitHub repository. This repo, currentai-org/os-ai-map, contains 1,184 YAML files detailing the curated projects, along with the schemas and scripts used for data gathering. More importantly for broad analysis, it includes a CSV file listing over 16,000 GitHub repositories the project is tracking. This transforms the map from a static report into a dynamic, queryable database for anyone to use.

How to explore the data

Because the data is hosted publicly on GitHub, it can be analyzed immediately without local setup. As Simon Willison demonstrates, the CSV of 16,185 repositories can be loaded directly into Datasette Lite via a URL. This allows for instant sorting, filtering, and faceting on metrics like star count, project language, or last push date. This accessibility is a key feature, lowering the barrier to entry for performing market analysis.

What's Interesting / What's Not

The most interesting aspect is the project's stated goal: to map gaps in the ecosystem. This framing elevates it from a simple catalog to an analytical tool. The project, backed by a $400 million non-profit, is positioned for longevity, suggesting it will be a consistently maintained resource rather than a one-off effort that becomes stale. The decision to release all underlying data under an MIT license is the correct one and represents the project's core value. It allows anyone to build their own analysis on top of Current AI's curation efforts.

The web visualization itself is less compelling. It's a fine entry point, but it doesn't expose the power of the underlying dataset. The mention of a "long tail" of 24,400 other uncategorized artifacts is currently just a number. Until those are categorized or scored, their value remains potential rather than actual. The initial curated list of 421 projects is a small slice of the total ecosystem, so its comprehensiveness is limited in this v0.1 release.

Pricing

Free (as of July 2026). The Open Source AI Gap Map and all its underlying data are provided by the non-profit Current AI. The data is licensed for use under the MIT license.

Verdict

The Open Source AI Gap Map is an essential new resource for anyone building or investing in the AI space. It provides a foundational dataset for strategic decision-making. If you are a founder choosing a tech stack or an investor performing market analysis, the os-ai-map repository should become a primary data source. Its true power is not in the polished UI but in the raw, queryable CSV and YAML files. For those willing to dig into the data, it offers a significant advantage in navigating the complex and fragmented open source AI landscape.

What We'd Test Next

For a v2 review, we would track the project's update cadence and its progress on categorizing the thousands of projects in its backlog. We would run specific, targeted queries to identify concrete examples of market gaps, such as under-resourced categories or emerging tools with high growth in stars but low overall visibility. We would also compare the comprehensiveness of its repository list against other well-known "awesome list" collections. Finally, once Current AI implements its scoring mechanism, we would evaluate its criteria and usefulness for identifying high-quality projects.

The investor read

The existence of a $400M non-profit to map the open source AI ecosystem signals the market's fragmentation and strategic importance. For investors, the Gap Map is not a product but a free intelligence service. The raw data in the GitHub repository is a quantitative starting point for market analysis, due diligence, and sourcing. Analysts can use it to identify crowded categories to avoid, find nascent projects with high velocity before they're widely known, and map potential gaps for new investments. While Current AI is a non-profit, the commercial opportunities lie in the gaps it illuminates. This dataset effectively commoditizes the initial discovery phase of landscape analysis, allowing funds to focus on deeper evaluation.

Pull quote: “The bottom line is that the Open Source AI Gap Map's primary value is its machine-readable, MIT-licensed dataset, not its web-based visualization.”