1 / 3
【Overview】 Kit for AI is a conversion and retrieval layer for AI agents. It takes documents, web pages and videos, turns them into clean Markdown or typed JSON, and then serves that content back to an agent as a searchable, citable knowledge base or as persistent memory. Agents reach it through MCP tools or a REST API, so the plumbing sits outside your application code. The problem it targets is the gap between raw files and usable context. PDFs, spreadsheets, slide decks and web pages arrive full of navigation chrome, layout noise and token overhead, and an agent that starts every session from zero forgets the user, the preferences and the decisions made last time. Kit for AI normalizes the input first — measured at up to 90% fewer tokens on real workloads — and then keeps it retrievable, with answers that cite the exact passages they came from. It is built for developers and ML engineers shipping AI agents, for teams assembling RAG pipelines, for builders who want persistent memory across Claude, Cursor, Windsurf, ChatGPT, Gemini and other AI tools, and for consultants and agencies that need per-client project isolation. A free tier covers a first prototype; paid plans scale the message, conversion and knowledge-base volume. 【Key features】 • Document conversion: Turns PDF, Word, Excel, PowerPoint, CSV, HTML and images into clean Markdown for LLMs and RAG. • URL and web to Markdown: Fetches a page and returns just the main content, stripping nav bars, ads and footers. • JSON extraction: Returns typed, structured data shaped by your own JSON schema instead of Markdown. • Knowledge bases and cited chat: Groups converted documents into named knowledge bases an agent can read and chat over, with answers citing the exact passages. • Persistent memory: Exposes remember, recall and search as MCP tools so an agent keeps users, preferences and decisions across sessions. • Hybrid semantic search: Runs vector plus full-text retrieval with reranking and returns the most relevant passages with source and score. 【Use cases】 1. RAG pipelines that need clean, chunk-ready Markdown for vector stores and retrieval. 2. Feeding AI agents structured context instead of raw PDFs. 3. Invoice and form extraction into typed JSON fields using your own schema. 【Pricing】 Freemium: $0/month — Chat with your documents at 20 messages a month, turn 10 files or web pages into clean text, turn 5 videos into searchable notes, and keep it all in 1 knowledge base. Pro: $9/month — 1,500 chat messages a month, 2,000 files or web pages, 150 videos turned into searchable notes, 25 knowledge bases and 25 GB of storage. Business: $29/month — Everything in Pro, with 10,000 messages a month, 20,000 files or web pages and 1,000 videos turned into searchable notes. Enterprise: Custom — Custom volume and SLAs, dedicated infrastructure, SSO and custom integrations, white-glove onboarding. 【Pros and cons】 Pros: - The value is in the wiring you skip: parsing, chunking and retrieval arrive as callable tools rather than as a stack you assemble and maintain yourself. - The free tier is generous enough to prove a RAG prototype on real documents before committing to a paid plan. - Named knowledge bases give agencies and consultants a natural unit of separation for per-client work. - Citing exact passages makes the output easier to defend to stakeholders than an unsourced chat answer. Cons: - JSON schema extraction is a Pro feature, so the free tier only returns Markdown. - deepExtract, the vision-based cleanup and figure description pass, is limited to Pro plans. - Bulk requests accept at most 25 file and/or URL fields per call. - Knowledge base counts are capped by tier: 1 on Free, 25 on Pro and 1,000 on Business. 【FAQ】 Q: How much does Kit for AI cost? A: There is a free tier at $0/month, Pro at $9/month and Business at $29/month, with custom Enterprise pricing for larger volume, SLAs and dedicated infrastructure. Q: Who is Kit for AI built for? A: Developers and ML engineers building AI agents, teams assembling RAG pipelines, builders who want persistent memory across AI tools, and consultants or agencies that need per-client project isolation. Q: Does Kit for AI work with the AI tools I already use? A: Yes. Knowledge bases and persistent memory are exposed as MCP tools or a REST API, so agents in Claude, Cursor, Windsurf, ChatGPT, Gemini and similar tools can call them directly.
Freemium · $0/month (Chat with your documents — 20 messages a month, Turn 10 files or web pages into clean text, Turn 5 videos into searchable notes) Pro · $9/month (Chat with your documents — 1,500 messages a month, Convert 2,000 files or web pages, Turn 150 videos into searchable notes) Business · $29/month (Everything in Pro, with room to grow, Chat with your documents — 10,000 messages a month, Convert 20,000 files or web pages) Enterprise · Custom (Custom volume + SLAs, Dedicated infrastruc
WPS AI is an AI assistant built into WPS Office that helps users write documents, generate presentation outlines and…
Baidu Comate is an AI coding assistant built by Baidu on its ERNIE model. It goes beyond simple autocomplete: it…
UniConverter is an AI media conversion and enhancement tool for video, audio, and images. It combines everyday…
Added
10/22/2026
Views
618
Clicks
4
6 similar tools to consider instead of Kit for AI