Minicod

What is Minicod?

Minicod is an academic data platform that puts Semantic Scholar, OpenAlex, and PubMed behind a single API. Developers can search more than 450 million papers, follow citation trails, and look up authors. An OpenAI-compatible chat endpoint answers research questions with real papers and verifiable DOIs, so the references are never invented.

Top Features:

• Unified paper search: query three scholarly sources through one key and one format.
• Grounded chat endpoint: ask plain-language questions and get papers with real DOIs.
• Citation data: trace references, citing papers, author h-index, and co-author networks.

Use Cases:

• Literature reviews: pull high-recall paper lists for reviews without juggling separate databases.
• Research assistants: give LLM apps real sources so answers cite existing papers.
• Journal selection: compare rankings, timelines, and submission resources before sending a manuscript.

Who Can Use Minicod?

• Developers: build academic search features and AI agents on one scholarly API.
• Researchers: find and verify sources without hopping between separate search engines.
• Systematic reviewers: run broad, repeatable searches that return identifiers for every record.

Pricing

• Free ($0): 1,000 API calls a month and 100 per day.
• Basic ($49 monthly): 50,000 calls a month, higher rate limits, and priority support.
• Pro ($199 monthly): 500,000 monthly calls, top rate limits, and an SLA guarantee.

Pros and Cons

Pros:

• No invented references: every result links to a real record with source identifiers.
• Generous free tier: a thousand monthly calls covers a small prototype.
• Agent friendly: the open-source MCP server connects paper search to agent clients.

Cons:

• Developer focused: the API is the main product, so nontechnical users may struggle.
• Uneven metadata: optional fields like abstracts depend on the upstream source.
• Paid plan jump: the first paid tier costs forty-nine dollars a month.

FAQs:

1) Which databases does it cover?

It combines Semantic Scholar, OpenAlex, and PubMed, covering over 450 million papers.

2) Does the chat endpoint work with OpenAI code?

Yes, it follows the OpenAI chat completions format, so existing clients work.

3) Is there a free plan?

Yes, 1,000 API calls each month are free, with no credit card needed.

4) Can AI agents use it directly?

Yes, an open-source MCP server lets agent clients search papers and journals.

5) What happens when my quota runs out?

Purchased call packages cover extra usage, and requests stop once both run out.