For researchers, PhDs & lab teams · Mac, iPhone, iPad

Private AI for academic research. Your unpublished work stays yours.

A private knowledge workspace for your papers, preprints, peer-review materials, lab notes, and reference libraries. Indexing stays local on your Mac. Cloud AI requests route through a zero-retention proxy, and on-device redaction runs before anything leaves.

Local-first indexing
Zero-retention proxy
On-device redaction
Offline mode

No credit card required · GPT-5, Claude & Gemini · Offline via Ollama

When your research involves

Unpublished drafts Preprints Peer-review reports Lab notes Embargoed data Grant proposals PDF reference library Interview transcripts

This isn't the kind of work you casually paste into ChatGPT.

Getting scooped or violating embargo is a real cost.

AI tools for research without sending drafts to OpenAI

The risk in research isn't HIPAA. It's getting scooped or breaking embargo. Elephas keeps indexing local and runs Smart Redaction before any cloud call, so you get GPT-5 quality without leaked drafts.

The best AI for research papers and literature review

Drop your reading list into a Super Brain: PDFs from arXiv, JSTOR, PubMed, plus your own drafts and notes. Ask for consensus methods, contradictions, and gaps. Citations point back to the page.

Built for deep research on your own corpus

Search at depth across papers, datasets, and field notes you trust, not the open web. Find every mention of a specific assay across 200 papers, or every transcript where a participant raised a theme.

An alternative to Elicit and Consensus for IP-sensitive work

Elicit and Consensus search the published literature. Elephas lets you reason over your own unpublished material with the same model quality, without putting it on a third-party server. Use both when the workflow demands it.

How Elephas works with your research

Indexing stays on your Mac. Redaction runs on-device. You choose whether inference happens locally, through the zero-retention proxy, or on your own API keys.

1

Drop your papers and drafts into a Super Brain

Create a Super Brain per project, dissertation, or paper. Add PDFs, your own drafts, lab notes, transcripts, and figures. Indexed entirely on your Mac.

2

Elephas indexes locally and redacts on-device

Common personal information, participant identifiers, and other structured identifiers are replaced with placeholder tokens before anything reaches a cloud model. Your full drafts and most metadata are never sent.

3

Pick the mode that fits the work

Run offline for embargoed data or pre-submission drafts, use cloud AI through the zero-retention proxy for general writing and synthesis, or bring your own API key to take the proxy out of the data path entirely.

Elephas data flow diagram: local indexing on your Mac, optional cloud AI with only relevant excerpts, on-device redaction before cloud processing

Three ways to run AI. You pick the mode for the task.

Indexing is always local. Inference is your choice, per Super Brain.

Offline AI

Local model on your Mac

Use a local model via Ollama. Both indexing and inference stay on your device. Nothing leaves your Mac.

When to use it

Embargoed datasets, pre-submission drafts, IRB-restricted transcripts, or any work you cannot risk having logged.

Cloud via zero-retention proxy

Default cloud mode

Requests route through an Elephas reverse proxy to providers configured with zero data retention. Providers do not log, store, or train on the content. Smart Redaction anonymizes common personal information on-device before the request is sent.

When to use it

Literature synthesis, methods comparison, writing assistance on non-confidential sections, general research questions.

Bring Your Own Key

Your own AI account

Connect your own OpenAI, Anthropic, or Google API key. Requests go straight from your Mac to the provider with no intermediary. Retention follows the policy on your own account.

When to use it

Lab teams with enterprise terms, or researchers who want to remove the Elephas proxy from the data path entirely.

Smart Redaction

Protection before processing, for sensitive research

When you use cloud AI, Smart Redaction scans outgoing queries and retrieved passages on your Mac and replaces common personal information with placeholder tokens before anything reaches the provider.

Coverage is strongest when sensitive details are common personal information or structured identifiers with clear patterns or labels. It is designed to reduce exposure, not to guarantee complete detection. For embargoed work, pair it with offline mode.

Read the full coverage and caveats

What's covered for research work

  • Participant names, contacts, and addresses
  • Subject IDs and study codes when labeled
  • Co-author and institution identifiers in drafts under review
  • Email addresses and phone numbers in interview transcripts
  • Links to internal data repositories and unpublished resources

Honest note: pattern matching catches common, structured identifiers. It does not guarantee every sensitive sentence in free-form narrative will be redacted. For embargoed or IRB-restricted material, run offline.

What it looks like in practice

Four workflows researchers run in Elephas every week.

Synthesize a literature review across 150 PDFs

Drop your reading list into a Super Brain. Ask for the consensus methods, contradictions between studies, and the gaps your paper could fill. Citations point back to the page.

Edit an unpublished draft without sending it to ChatGPT

Load your draft into a private workspace, run editing prompts locally with Ollama, or via the zero-retention proxy for non-confidential sections. Your manuscript does not enter a public training set.

Find the right paper, fast, across your reference library

Ask in plain English: "Which paper used the modified assay on stage IV samples?" Elephas searches the local index and cites the paper and page.

Code interview transcripts for recurring themes

Load qualitative transcripts into a project Super Brain. Ask for every passage where a participant raised a given theme, with the source line cited.

Elephas vs Elicit, Consensus, NotebookLM, and ChatGPT

Different tools, different goals. Here is what fits where for research work.

Elephas Elicit Consensus NotebookLM
Primary focus Your own corpus and drafts Search the published literature Search the published literature Your uploaded sources via Google
Mac-native app Yes Web Web Web
Indexing stays on your device Yes No No No
On-device redaction before cloud LLM call Yes N/A N/A No
Works fully offline with a local model Yes No No No
Bring your own OpenAI or Anthropic key Yes No No No

Comparison reflects published feature sets as of 2026. Use Elicit or Consensus for searching the open literature. Use Elephas for reasoning over your own work.

Not another chatbot

Generic AI tools were built for open prompts on the open web. Your research is neither.

Generic AI chat

  • × Consumer prompts may be used to train future models.
  • × Confidently invents a reference, and you have to check every citation.
  • × Cannot see your actual reading list or drafts. You paste in excerpts and lose context.

Elephas

  • Default cloud requests route to providers with zero data retention. Your drafts are not training data.
  • Answers cite the paragraph from a paper in your library, so you can verify in one click.
  • Indexed locally. Elephas works against your reference library without uploading it.

Questions researchers ask before they install

Straight answers, with links to the underlying docs.

  • Consumer ChatGPT may use your inputs to improve future models. For pre-publication drafts, peer-review materials, or unpublished data, that is a real exposure. Elephas routes cloud requests through a zero-retention proxy and runs Smart Redaction on your Mac first, so your drafts and IP do not become training data. For embargoed work or anything you cannot risk leaking, run a local model and nothing leaves the device.
  • Elicit and Consensus search the published literature and summarize papers from the open corpus. NotebookLM lets you load your own sources but routes everything through Google. Elephas is local-first: your PDFs are indexed on your Mac, cloud requests go through a zero-retention proxy with on-device redaction, and you can run a local model offline. Different goal: keep your unpublished work private while still using GPT-5 or Claude on it.
  • No. Elephas does not train any models on your data. Cloud requests through the Elephas proxy only go to providers with zero data retention enabled, so providers do not log, store, or train on your question or the passages sent with it. With Bring Your Own Key, requests go directly to the provider on your own account.
  • Yes. Drop PDFs into a Super Brain and Elephas indexes them locally on your Mac. You can mix downloaded papers, your own drafts, peer-review reports, notes, and figures in the same workspace. Search across the whole corpus and get answers cited back to the page.
  • Yes. Bring Your Own Key sends requests directly from your Mac to the provider using your own API account, so retention follows the policy on that account. This removes the Elephas proxy from the data path entirely. Configure it in Settings under AI Provider.

Unpublished work deserves a better default

Use AI on your research with more control and less exposure

Local-first indexing, on-device redaction, and a zero-retention proxy by default. Pick the mode that fits the work.

Lab or department evaluation? Contact us.