Memory · Resource

Documents, indexed and recallable.

Resource memory holds every artifact your agent touches — PDFs, spreadsheets, models, blobs — with Tantivy-backed full-text search and content-addressed deduplication.

Start free Read the docs

Capabilities

What it does.

Every capability is production-grade. No flags, no betas — these ship the day you adopt Minds.

Tantivy search

BM25 + fuzzy + faceted queries across every blob.

Content-addressed

BLAKE3 hashing dedupes identical content automatically.

Streaming

Multi-GB files stream — no full-load into RAM.

Format-aware

PDF, DOCX, MD, CSV, Parquet parsed with structure preserved.

Versioned

Every update produces a new version; old versions remain queryable.

Model store

Store ONNX/Candle/PyTorch models as first-class resources.

API

A few lines is all it takes.

The SDK reflects the conceptual model directly. No glue code, no orchestrator, no learning curve beyond the data shape.

exampletypescript

await minds.memory.resource.upload({
  path: "./Q3-board-memo.pdf",
  tags: ["finance", "board", "Q3-2026"],
});

const matches = await minds.memory.resource.search({
  query: "channel mix changes",
  tags: ["finance"],
  topK: 10,
});

Specs

What it costs, what it guarantees.

Performance

Max blob size: 50 GB (streaming)
Indexing rate: ~120 MB/sec per node
Full-text query: <8ms P99 over 1M docs
Dedup: BLAKE3 content addressing

Guarantees

Durability: fsync per WAL commit
Isolation: MVCC · snapshot
Audit: BLAKE3 Merkle chain
Encryption: AES-256-GCM per-namespace
Concurrency: 100K+ ops/sec

Keep exploring

Vector + SQL

Run analytics over indexed content.

Knowledge Graph

Resources link into the graph.

Vault

Encrypted store for sensitive resources.

Build it on Minds.

Start with the SDK. Ship in an afternoon. Run anywhere — Akasha Cloud, on-prem, or air-gapped.

Start free View pricing