Infrabase.ai
Andersson-Larsson Holding AB · Business · Other
Infrabase.ai is a hand-verified directory of AI infrastructure tools, from inference APIs and GPU clouds to vector databases and agent frameworks, with a free JSON API.
About Infrabase.ai
What Is Infrabase.ai
Infrabase.ai is an AI infrastructure directory. It covers the parts you assemble around a model: inference APIs, GPU clouds, vector databases, agent frameworks, LLM observability tools, fine-tuning platforms, and audio APIs. The site currently lists 356 products, and a person reads every one of them before it goes up.
Each listing carries pricing, headquarters country, GDPR status, licence, and GitHub stars alongside a short write-up for developers. Those fields get checked against the vendor's own site rather than against search results or another directory. When a detail can't be confirmed, it's left out. That's a slower way to build a directory, and the result is smaller than most. The data is more useful for it.
The site takes no position on winners. There is no best field and no editorial ordering, because which tool fits depends on constraints the site can't see. Comparison pages line up sourced numbers and stop there. It's operated by Andersson-Larsson Holding AB in Sweden, and one developer runs it day to day.
The limits are real. 356 products is a narrow slice of a crowded market, so popular options can be missing simply because nobody submitted them. Prices change between reviews, and corrections wait on one person's inbox.
Getting Started
- Open infrabase.ai and choose a category from the directory, such as inference APIs, vector databases, or agents.
- Narrow the list by job, GDPR status, or HQ country. The job filter says what a product actually does, which is more precise than the category name.
- Open a product page to read the write-up and check pricing, hosting country, and licence details.
- Put two or three candidates on a shortlist with the alternatives and comparison pages.
- For programmatic access, call the AI infrastructure API at /api/query or connect the MCP server to your coding assistant.
Product Information
A quick look at Infrabase.ai's pricing, supported platforms, and performance.
Best for
The users, tasks, and scenarios where this tool fits best.
Users
- Backend and ML developers
- Technical founders
- Compliance and procurement teams
Tasks
- Choosing an inference API
- Shortlisting vector databases
- Checking whether a vendor can take EU data
- Feeding an AI agent structured tool data
Scenarios
- The week before a build starts, when the stack is still a list of candidates and every choice locks in a migration later.
- A vendor review where someone asks for the hosting country and data terms in writing, and you need a second source that isn't the sales deck.
- Inside an IDE assistant, when a developer asks which LLM observability tools support a specific framework and expects an answer with links.
Key features
Hand-Written Listings
Every product page is researched by a person who opens the vendor's site and checks the claims. The write-up covers what the tool does, what it costs, and where it breaks down. That matters most in categories where homepages describe the same capabilities in nearly identical words.
Structured Fields You Can Filter
Listings carry pricing, HQ country, GDPR status, licence, primary job, and GitHub stars. You can filter on those fields directly instead of reading ten pages to find out which vendors are based in the EU. The same fields come back in every API response, so you can sort and compare without scraping.
Job-Based Search
Categories are broad buckets, and a lot of tools sit awkwardly inside them. Infrabase tags each product with the job it does, such as hosted inference, GPU compute, or agent hosting. Search by job and you get tools that solve your problem rather than tools that share a label. A team picking an AI agent framework sees hosting platforms kept apart from orchestration libraries.
Open JSON API
The AI infrastructure API answers GET requests at /api/query with no key and no signup. You can search by text, category, job, GDPR status, or country, then page through results with an offset. Responses are cached for about an hour, and every result includes the date the listing was last checked.
MCP Server for Coding Agents
The site runs an MCP server over Streamable HTTP, also without a key, so Claude Code, Cursor, or a similar client can query the directory mid-task. It exposes the same data as the HTTP endpoint. Claude users can also install it from Anthropic's connector directory.
Price Tables and Comparison Pages
A separate price endpoint returns per-model token prices read off each provider's pricing page, with the source URL and the date it was verified. Comparison pages put those numbers side by side. Neither ranks the products, which keeps the page useful as a reference rather than a recommendation.
Pros and cons
Pros
- Fields are checked against the vendor's own site, so pricing and hosting details hold up better than on scraped directories.
- The API and MCP server are free, need no account, and publish under CC-BY-4.0, which makes the data easy to reuse in internal tools.
- Job filters answer "what does this do" faster than category browsing.
- No pay-to-rank: featured placement is labeled, and paid slots don't change the comparison order.
- Submitting a product costs nothing, and every page has an edit link for corrections.
Cons
- Coverage stops at roughly 356 products, so a missing tool often means nobody submitted it, not that it was rejected.
- Pricing fields are snapshots, which means a quote from three months ago can be wrong by the time you sign up.
- There are no user reviews, ratings, or usage numbers, so community sentiment has to come from somewhere else.
- One person runs the site, so corrections and new categories move slower than a staffed directory would.
Frequently asked questions
Yes. Browsing the directory, querying the API, and submitting a product all cost nothing. The only paid item is a featured listing for vendors, at $49 per month or $490 per year.
Related content
Explore related tools, skills, and articles for Infrabase.ai.
Infrabase.ai Alternatives
Knocket
Knocket (Tencent RTC team) · Productivity · Business · ChatbotKnocket is a free live chat and contact page tool that lets visitors message you from your website, app, or a shareable link, with every reply handled from Telegram or one inbox.
shippedit.me
Krawlup · Marketing · Businessshippedit.me now redirects to Krawlup, a web platform that audits SEO, AEO and page speed every week, then hands the fixes to AI coding agents over MCP.
Ayven
Alknoma · Productivity · BusinessAyven is an AI control centre by Alknoma that turns a plain-language goal into a dated plan, hands the work to your people and AI agents, and asks for your approval only when the call is yours.