AgentIndex icon
AgentIndex
Wax logo
Wax
★ 805
vs
haiku.rag logo
haiku.rag
★ 634

Wax vs haiku.rag

Wax: Wax provides persistent, searchable, and private on-device memory for iOS and macOS AI agents, eliminating the need for servers or cloud services. It integrates document storage, semantic search (HNSW), full-text search (BM25), and token budgeting into a single, crash-safe .wax file.; haiku.rag: Haiku RAG is an agentic retrieval-augmented generation (RAG) system built with LanceDB, Pydantic AI, and Docling. It offers advanced features like hybrid search, multi-agent workflows for research, complex analytical tasks via code execution, and conversational RAG with session memory.

01

TL;DR

Wax logo
WaxRAG / Knowledge Base · ★ 805

Conversational agents remembering user preferences

Read Wax full review →
haiku.rag logo
haiku.ragRAG / Knowledge Base · ★ 634

Indexing various document types (e.g., PDFs, URLs) for search and retrieval.

Read haiku.rag full review →
02

Side-by-Side Comparison

Field
Wax logoWax
haiku.rag logohaiku.rag
Category
RAG / Knowledge Base
RAG / Knowledge Base
Stars
★ 805
★ 634
License
Apache-2.0
MIT
Install Command
Add to Package.swift dependencies: .package(url: "https://github.com/christopherkarani/Wax.git", from: "0.1.8")
pip install haiku.rag
Updated
3w ago
1d ago
Open Source
Yes
Yes
Website
↗ Visit
↗ Visit
GitHub
↗ GitHub
↗ GitHub
Tags
AI Agents, On-device AI, Vector Database
RAG, Agentic AI, Hybrid Search
03

Features

Wax logoWax
01Hybrid retrieval (BM25 + HNSW)
02On-device embeddings (MiniLM)
03Metal acceleration for GPU
04Automatic token budgeting
05Single portable .wax file
haiku.rag logohaiku.rag
01Hybrid search — Vector + full-text with Reciprocal Rank Fusion
02Research agents — Multi-agent workflows via pydantic-graph: plan, search, evaluate, synthesize
03RLM agent — Complex analytical tasks via sandboxed Python code execution
04Conversational RAG — Chat TUI and web application for multi-turn conversations with session memory
05Local-first — Embedded LanceDB, no servers required. Also supports S3, GCS, Azure, and LanceDB Cloud
04

Use Cases

Wax logoWax
↳Conversational agents remembering user preferences
↳Note-taking apps with semantic search
↳Personal assistants learning habits privately
↳On-device RAG pipelines for sensitive data
haiku.rag logohaiku.rag
↳Indexing various document types (e.g., PDFs, URLs) for search and retrieval.
↳Performing question answering with source citations to verify information.
↳Executing complex analytical tasks using sandboxed Python code (RLM mode).
↳Engaging in multi-turn conversations with memory for contextual understanding.
↳Integrating as a tool for AI assistants (e.g., Claude Desktop) for document management, search, and QA.
FAQ

FAQ

What is the difference between Wax and haiku.rag?
Both Wax and haiku.rag are in the RAG / Knowledge Base category. Wax has 805 stars, while haiku.rag has 634 stars.
Which is better, Wax or haiku.rag?
The best choice depends on your use case. Choose Wax if Conversational agents remembering user preferences, and haiku.rag if Indexing various document types (e.g., PDFs, URLs) for search and retrieval..
Is Wax free or open source?
Yes, Wax is open source on GitHub (Apache-2.0).
Is haiku.rag free or open source?
Yes, haiku.rag is open source on GitHub (MIT).
Cloud Infrastructure
Ad

Deploy on DigitalOcean

Host your MCP servers, Docker containers, and AI agents with $200 free credit over 60 days.

Claim $200 Free Credit
→

Related

Alternatives to Wax →Alternatives to haiku.rag →Wax details →haiku.rag details →