Kimi alternatives for chat and document work

AnythingLLM suits replacing Kimi document work with workspaces that return source citations. The desktop app serves personal files, while Docker adds shared access. Agents, scheduled tasks and workspace memory extend the file conversation with local or hosted models.

Cherry Studio covers Kimi-style file questions and reusable chat assistants in a desktop app. Work with PDFs, Office documents or images and compare conversations with different models. MCP connections add external tools; inference uses your chosen local backend or hosted provider.

RAGFlow answers questions across office files, scans and structured data with traceable supporting references. Multi-step retrieval can gather more context for complex questions, and you can inspect the extracted document chunks.

LibreChat adds a self-hosted workspace for Kimi-style file conversations and agent tasks. It supports searchable conversations, tools and code artifacts, with multiple model providers behind one interface.

kotaemon focuses on the document-research part of Kimi. Ask across multiple files and inspect citations in a PDF viewer, including highlighted passages, with local models and document parsers available.

WeKnora combines cited document answers with agent tasks over the same knowledge bases. Office files and connected sources can feed a linked wiki, while MCP tools and skills let the agent act on the material it retrieves.

DocsGPT suits the company-document side of Kimi. Search files and connected knowledge, then check source titles and page numbers in the answer. Its self-hosted platform adds agent workflows and team access controls, with local or hosted model backends.

Open WebUI covers Kimi conversations and document retrieval in a shared chat workspace. Attach notes to chats, combine model instructions with knowledge and tools, and manage user access. Use local backends or hosted providers rather than Kimi membership access.

RecurseChat provides file conversations over PDFs, Markdown and text files on a Mac. On-device embeddings and vector search retrieve material for answers, with a floating window for keeping document questions beside your other work.

LM Studio suits the personal-file part of Kimi in a desktop app. Download a model, ask questions about documents on your computer and use its local API for other applications. It focuses on local model management rather than Kimi's hosted service.

QAnything suits Chinese and English document collections, including questions asked in a different language from the files. It retrieves and reranks passages from office documents, email and other sources before answering.

Yuxi covers cited questions across documents and spreadsheets in a shared agent workspace. Agents can analyze the material and deliver files for preview or download, while knowledge graphs connect facts across the source documents.

localGPT organizes document collections and topic-based chats for questions about your files. Answers include source passages you can inspect, and combined keyword and semantic search helps find both exact wording and related material.