Writing and content decision
Jenni AI
A minimal research-writing tool (personal library + citation-grounded autocomplete) is feasible for a small team to build and run, but Jenni's claimed large proprietary index and scale features are hard to replicate, so building is viable only for a narrower workflow.
Visit website↗Open-source builds that already do this
Every project below is open source and already does this job today. Fork one, self-host it, or take the parts you need — the build prompt further down assumes an empty file, and this is the shortcut past that. Licences differ; check the one on each card before you ship.
What a replacement has to do
- Provide an editor where users upload PDFs or connect a Zotero/Mendeley export, index the docs into a vector DB, offer semantic search over that index, generate AI autocomplete responses grounded to specific PDF pages (with a link to page/paragraph), and export documents with inline citations.
What it still won’t have
- Access to Jenni's claimed 200M+ indexed papers
- Out-of-the-box traceability to a large proprietary academic index
- Unlimited/autoscaling AI usage and institutional/team features
- Priority support and other paid-plan niceties
What remains hard
- Proprietary data
let Jenni search 200M+ papers .
First-year cost
Keep paying
Paying is—cheaper in year one.
On cash alone, building overtakes the subscription at 14 seats.
Money you would actually spend
Time you would spend
—
What you would spend
What we assumed
The verdict above measures whether you could build it. This one is only about money.
Runnable build prompt
Build a minimal self-hosted research-writing workspace using Next.js for the frontend, Node/Express backend, PostgreSQL for user/docs metadata, Pinecone (or open-source vector DB) for embeddings, and OpenAI (or similar) for embeddings + text generation. In scope: PDF upload + text extraction (per-page), embedding pipeline, per-page vector index, semantic search API returning page/paragraph offsets, an editor with AI autocomplete that conditions on retrieved passages and inserts inline citations (page numbers and source filename), and DOCX/LaTeX export. Out of scope: indexing a 200M-paper corpus, institutional billing, team admin console, and priority support. Include error handling for failed uploads, rate limits, and corrupted PDFs, and provide unit/integration tests for the upload, index, search, and generation flows.
How we checked
How the score was reached
- Partly verdict base52
- An open-source build was found+5
- 3 cited sources+3
- Price verified on pricing page+3
- 3/3 assessment runs agreed+4
- Hard moats found in the evidence-3
- Evidence score64
The base comes from the verdict. Everything under it is a check that either happened or did not, and each one is a fact frozen in this record rather than a judgement made at render time — so the same evidence always produces the same number.
How scoring works →Cited sources · 3
Every page the run actually retrieved.
- official productJenni | AI Academic Writer & Research Tool for Students & Academics
- official pricingJenni AI Pricing & Plans | Find the Right Plan for Your Writing
- open sourcedixiyao/LLM-Academic-Writing
Integrity checks
What held up, and what did not.
