On-device · No cloud · No tracking

Private AI
for your documents.

Chat with any PDF or Word document — fully offline. Anvit reads, reasons over, and answers questions about your files without sending a single byte to the cloud.

FreeWorks in airplane modePowered by Gemma 4
100% OfflineNo AccountsCited Sources
Nothing uploaded0 bytes leave your phone
Agentic RAGReasons in multiple steps
Anvit home screen — Ask Anvit anything
100% on-device
Every answer is computed locally
0 bytes
Uploaded to any server, ever
Gemma 4
Google's open model, on your phone
No sign-up
No accounts, no analytics, no telemetry
Agentic RAG

It doesn't just search.
It reasons.

Plan→ Decompose→ Search→ Filter→ Answer

Complex questions are automatically broken into focused sub-queries, retrieved across multiple passes, and refined before you ever see the final answer.

  • Hybrid retrieval — dense semantic search + BM25 keyword search find what you mean, not just what you type.
  • Corrective RAG — when retrieval quality is low, Anvit re-queries instead of hallucinating.
Agentic reasoning steps
Think mode — chain of thought
Think Mode

See the model think.

Toggle Think Mode to watch Gemma's full chain-of-thought as it analyzes your question — how it reads the request, scans for keywords, and weighs each source before answering.

  • Transparent reasoning on demand — nothing hidden behind a black box.
  • Trust what you read — follow the model's logic step by step.
Cited Sources

Every answer cites
its source.

No hidden hallucinations. Each response links back to the exact document and the passages it was built from — so you can verify everything in a tap.

  • Verifiable — open the source passage behind any claim.
  • Traceable — every answer carries a reference and timestamp.
Cited sources for every answer
Document collections
Collections

Organize documents into
smart collections.

Group your PDFs and Word files into named libraries — Research, Work, Study — and scope each chat to exactly the right set. Anvit indexes everything into searchable chunks, ready for Q&A.

  • Per-collection chat — ask within one library or across them all.
  • Indexed locally — chunked and embedded entirely on your device.
Offline & Background

Works offline.
Runs in the background.

Drop into airplane mode and Anvit keeps working — no connection required after the model downloads. Long jobs like ingestion and inference keep running even when you switch apps or lock your phone.

  • Airplane-mode ready — full functionality with zero signal.
  • Background tasks — indexing and answers continue while you're away.
Background indexing notification
Generating an answer in airplane mode
Under the hood

Powered by open AI.

An open demonstration that powerful AI no longer requires someone else's servers. Your phone is enough.

Language Model

Google Gemma 4

Runs fully on-device via Google LiteRT-LM, on CPU or GPU.

E2B
1.3 GB
E4B
2.5 GB
Embeddings

EmbeddingGemma · Gecko

High-quality on-device embeddings power semantic document search.

EmbeddingGemma
300M
Gecko
110M
Retrieval

Agentic + Corrective RAG

Multi-pass reasoning with hybrid dense + BM25 retrieval and self-correction.

Search
Hybrid
Hallucinations
Guarded
Built With

Kotlin · Compose · M3

100% native Android — Jetpack Compose and Material 3, end to end.

Runtime
LiteRT-LM
Cloud calls
Zero
Privacy by design

Your documents never leave.

Privacy isn't a setting in Anvit — it's the architecture. There's no server to send anything to.

No data collection

Your files, questions and chat history stay on your device. Always.

No accounts

No sign-ups, no logins. Install and start asking immediately.

No internet needed

No connection required once the model is downloaded. Period.

Open about how it works

No analytics, no telemetry. Transparent reasoning you can inspect.

FAQ

Questions, answered.

Is Anvit really fully offline?+
Yes. After the AI model downloads once, Anvit needs no internet connection — it even works in airplane mode. No connection is required to use it.
Where do my documents and chats go?+
Nowhere. Your PDFs, Word files, questions and chat history never leave your device. There is no account, no server, and no analytics.
What file types can I chat with?+
Both PDF and Word (.docx) documents. Organize them into collections and ask questions across one library or all of them.
What AI model powers Anvit?+
Google's open Gemma 4 model (E2B or E4B) runs on-device via LiteRT-LM, with EmbeddingGemma or Gecko handling document search.
Can I trust the answers?+
Every answer cites the exact source passages it used, and Think Mode lets you watch the model reason step by step — so you can verify, not just trust.
How does Anvit read and index my documents?+
When you add a document, Anvit parses its structure — headings, sections, tables, and lists — then splits it into section-aware chunks of up to 8,000 tokens each. Tables are kept intact as structured chunks so data rows aren't mixed with prose. Repeated page furniture like page numbers and company footers is stripped out. Each chunk remembers the section heading it came from, so when you ask a question Anvit can retrieve the exact passage and tell you precisely where in the document it found the answer.
Is there an iOS version?+
Anvit is live on Android today. An iOS version is in active development — add your email below and you'll be the first to know when it launches.
Anvit

Chat with your documents.
Stay private.

Download Anvit and turn every PDF and Word file into searchable, conversational knowledge — without giving anything away.

Get it onGoogle Play