Features

Everything Cobentt offers. Nothing more.

A complete AI workspace with no feature bloat. Every tool has a purpose.

Core Workspace

Chat with any AI model from a single interface. Compare responses side by side. Switch models mid-conversation without losing context.

Organize everything into projects with folders, tags, and version history. Every chat, document, and research source belongs to a project.

Write, edit, and publish documents with AI collaboration. Live Markdown preview, export to PDF, DOCX, and LaTeX.

Meridian maintains a persistent knowledge graph across all your projects, chats, documents, and research. Context never resets.

Purpose-built SLM that plans, routes, retrieves context, selects models, and verifies every response before you see it.

Routes each request to the optimal model based on task complexity, capability requirements, cost, and latency.

Built-in content filtering, prompt injection detection, output validation, and data privacy controls.

Remembers your preferences, project history, and research context across sessions. Everything stays connected.

Multi-source research with PDF understanding, citation generation, source tracking, and knowledge synthesis.

Built-in browser for research, testing, and web automation. AI can browse, extract, and analyze web content.

Upload documents, PDFs, and notes to build a personal knowledge base. Search across all sources instantly.

Auto-capture citations, manage references, and export in any format. Supports APA, MLA, Chicago, and more.

Built-in code editor with syntax highlighting, AI pair programming, and multi-file context understanding.

Deploy autonomous agents that plan, reason, call tools, write code, and execute complex multi-step tasks.

Visual canvas for system design, architecture diagrams, brainstorming, and flowcharts.

Generate and edit images with DALL-E, Stable Diffusion, Flux, and more. All from within the workspace.

Voice input and output for hands-free interaction. Works with all supported models.

Extend Cobentt with plugins, custom tools, and third-party integrations via the extension API.

Run models locally with Ollama. Full support for local inference without cloud dependency.

Local-first architecture. Your API keys. No training on your data. End-to-end encryption.