agentic-ai-apis: A Curated Directory of APIs for Building Autonomous AI Agents
The ultimate collection of APIs for building autonomous AI agents — 2,036 production-ready APIs across Agents, AI Models, and MCP Servers. Stop wasting weeks building infrastructure. Plug these in and ship your agent today.
At a glance
- What is it?
- agentic-ai-apis is a GitHub repository containing a curated list of 2,446 production-ready APIs across three categories: Agents, AI Models, and MCP Servers, maintained through a daily-syncing GitHub Actions workflow.
- Who is it for?
- This repository is useful for developers who are starting an agentic AI project and want a pre-filtered starting point rather than raw API discovery across dozens of provider websites. It is not an SDK, not a framework, and not a code library.
- Can I use it commercially?
- Not without permission. GitHub finds no licence file in the repository, and without a licence all rights are reserved by default: you may read the code but not reuse it. Check the README, or ask the authors, before using it.
- Is it still maintained?
- Yes. The repository last received commits 2 days ago.
- What is it written in?
- Mainly JavaScript, according to GitHub's language statistics.
Answers come from the project's GitHub data, last synced on September 29, 2026, and from our analysis. They are not legal advice.
Editorial analysis
What This Repository Is and Who It Serves
Finding APIs for an agentic AI project is time-consuming because the relevant options are spread across dozens of provider websites under inconsistent naming. agentic-ai-apis consolidates 2,446 API listings into three categories, each scoped to a distinct layer of an autonomous AI system.
The project targets developers building autonomous assistants, AI copilots, MCP toolchains, research agents, or internal automation tools. The README describes the design intent as a launchpad, not a junk drawer, with tight scoping around the layers that matter for autonomous systems. Anything outside agents, AI models, and MCP servers has been removed from the tracked structure.
The repository does not contain code for calling these APIs. Each entry links to the provider's page. The value is in the curation and categorization: narrowing 2,446 candidates to a shortlist without reading every provider website individually.
Repository Structure and Three Categories
The repository has three primary directories, each containing its own category README.
The `agents-apis/` directory covers 699 APIs. The README describes this category as execution layers, orchestration, autonomous task handling, and agent-style workflows.
The `ai-models-apis/` directory covers 1,338 APIs. This is the largest category, covering generation, reasoning, extraction, transformation, and model-powered product building blocks.
The `mcp-servers-apis/` directory covers 409 APIs. This category covers Model Context Protocol integrations that connect assistants to real tools, systems, and data: docs, search, analytics, scheduling, and developer workflows.
The `settings/` directory contains the generation scripts that rebuild the three categories. A GitHub Actions workflow in `.github/` syncs the Apify catalog daily and commits only when upstream data actually changes, which means the repository history reflects real catalog updates rather than empty commits.
How to Use the Directory
The README gives a four-step process for using the repository. Pick the relevant category first: Agents for execution and orchestration, AI Models for inference and generation, or MCP Servers for tool integrations. Then open that category README and scan the API names and descriptions. Click through to the provider page for implementation details, pricing, and documentation. Build a shortlist rather than reading every entry.
The repository has no install step and no code to run. Cloning it is optional; the README and category directories can be browsed directly on GitHub. There is no CLI, no package, and no configuration format specific to this project. The entire interface is markdown files with links.
The README notes that the discovery is faster because the clutter has been removed. Whether that claim holds depends on what you compare it to: against a raw web search, a pre-filtered list of 2,446 categorized entries is faster; against a specialized API registry, the comparison depends on the registry's own quality.
Daily Sync and Data Provenance
The repository's content is not manually curated entry by entry. A GitHub Actions workflow syncs the Apify catalog daily. The README states that commits are made only when upstream data actually changes, preventing a noisy commit history from automated runs.
The generation scripts in `settings/` control which categories are rebuilt and how the visual README layout is generated. The README states that API links carry existing affiliate tracking from the upstream source data. This means clicking through to a provider page may credit the repository owner through an affiliate arrangement.
The README updated timestamp in the README itself shows 2026-09-27, and the last push to the repository was on 2026-09-26. The category counts (699 agents, 1,338 AI models, 409 MCP servers) add to 2,446, matching the headline figure. The description field in the repository uses the figure 2,036 but the README uses 2,446; the README is the more current source.
What the Repository Does Not Provide
agentic-ai-apis has no license. The repository page lists the license as unknown, which means there is no explicit grant to copy, redistribute, or build derivative works from the content. This is a meaningful constraint for teams that want to incorporate the directory data into their own tooling or documentation.
There is no validation of the listed APIs. The repository does not test endpoints, check uptime, or verify that listed capabilities match current provider documentation. An API listed as production-ready reflects the upstream source data, not an independent assessment.
The repository provides no integration code, no SDK wrappers, and no example applications. Teams that want to evaluate a specific API still need to go to the provider page and write their own integration. The repository is a filtered list, not a tested integration layer.
There is no search functionality within the repository itself. Finding a specific API requires scrolling or using GitHub's text search on the category README files.
How This Compares to awesome-lists
The awesome-list format on GitHub collects links and brief descriptions of tools in a given domain, maintained by community contributors through pull requests. agentic-ai-apis differs in two ways: its content is generated automatically from a data catalog rather than curated manually, and it is scoped to a single domain (agentic AI infrastructure) rather than a broad technology area.
The automated generation means additions and removals happen on the catalog's schedule, not on contributor availability. The narrow scope means the list stays on topic without the sprawl that affects some community-maintained awesome-lists. The tradeoff is that automated lists carry whatever the upstream source contains, including entries that might not survive human editorial review. Teams evaluating APIs from this list should still check each provider independently for current availability and pricing.
Licensing Gap and Affiliate Link Disclosure
Two facts about this repository deserve attention before relying on it heavily. First, it has no license. The repository lists the license as unknown. Using the content in a commercial product, building a tool that incorporates the data, or redistributing the list requires clarifying the terms with the maintainer, because the default copyright position without a license is that no use rights are granted beyond browsing.
Second, the README states directly that API links carry existing affiliate tracking from the upstream source data. Following a link from this repository to a provider's pricing or signup page may trigger affiliate attribution. This is disclosed in the README but easy to miss when browsing the category directories.
The last push to this repository was on 2026-09-26. The FOLLOW_CREATOR.md file at the repository root suggests the maintainer actively promotes the project.
Editorial conclusion
This repository is useful for developers who are starting an agentic AI project and want a pre-filtered starting point rather than raw API discovery across dozens of provider websites. It is not an SDK, not a framework, and not a code library. The 2,446 links point to provider pages, not to code or configuration. Before using it, check the license section: the repository has no stated license, which affects how you can incorporate its data into derivative works. Also note that API links carry affiliate tracking from the upstream source.
Frequently asked questions
What APIs does the agentic-ai-apis directory cover?
The repository covers 2,446 APIs in three categories: 699 Agents APIs for execution and orchestration, 1,338 AI Models APIs for generation and reasoning, and 409 MCP Servers APIs for tool integrations. Content is synced daily from the Apify catalog.
Is agentic-ai-apis a framework or an SDK?
No. It is a curated directory of API links organized into three markdown files. There is no code to install, no package to import, and no CLI. Every entry links to the provider's own documentation page.
Does agentic-ai-apis have a license?
No. The repository lists the license as unknown, which means no explicit use rights are granted. Teams that want to incorporate the directory data into their own products should clarify terms with the maintainer.
Official sources
Add this badge to your README
If you maintain this project, the badge below links readers to this analysis and shows its maintenance status from the daily GitHub snapshot. Paste the markdown into your README; add ?metric=license or ?metric=stars to the image URL for a different field.
[](https://hysenlabs.com/projects/cporter202-agentic-ai-apis)