Vela
FreeAI-powered IDE for novel writing — local LLM + RAG, privacy-first, BYOK. For web fiction authors and creative writers.
FreeFree tier
Inputs: textOutputs: text
About Vela
Vela is an open-source, privacy-first AI-powered IDE for novel writing, designed for web fiction authors, indie writers, and creative professionals. It combines a full AI writing pipeline (worldbuilding, auto outline, chapter drafting, rewrite, refine, review) with a local RAG knowledge base that handles millions of words. All data and model calls run locally using the user's own API keys (BYOK). The IDE features a resizable multi-panel layout, deep dark theme, keyboard shortcuts, and cross-platform support. It supports multiple LLM providers including OpenAI, DeepSeek, Gemini, Claude, Ollama, and GLM, and can route different tasks to different models. Vela also integrates the Model Context Protocol (MCP) for extensibility and includes built-in usage analytics.
Key Features
AI-powered novel writing pipeline: worldbuilding, auto outline generation, chapter drafting with streaming, intelligent rewrite, grammar and error refinement, automated plot review, and a triple post-process pipeline (Rewrite → Refine → Review) for high-quality output
Million-word local RAG knowledge base: bulk import of reference documents, semantic vector search for relevant context during writing, fully local storage via SQLite and lightweight vector engine (offline-capable)
Extensible architecture: BYOK support for OpenAI, DeepSeek, Gemini, Claude, Ollama, GLM with intelligent task routing; native MCP protocol integration for custom tool servers; built-in usage analytics dashboard
IDE-grade productivity UI: resizable multi-panel layout (file tree, editor, AI panel, terminal), immersive dark theme, customizable window management, global keyboard shortcuts, cross-platform (macOS, Windows, Linux)
Pros & Cons
Pros
- Privacy-first: all data stored locally, no cloud dependency for storage or processing (can use local models)
- Open source and free to use with no licensing fees
- Supports multiple LLM providers and allows intelligent routing of different writing tasks to different models
- Rich AI pipeline goes beyond simple chat to provide structured novel writing workflows
- Local RAG enables large knowledge bases without privacy concerns or API costs
- Extensible via MCP protocol for custom tool integration
- Cross-platform support (macOS, Windows, Linux) with IDE-grade UI features
Cons
- Requires technical setup: users need to obtain and configure their own API keys for cloud models or run a local LLM
- May have a learning curve for non-technical writers unfamiliar with API keys, local models, or IDE-like interfaces
- Community and documentation are still growing (GitHub stars indicate early-stage project)
- Limited out-of-the-box integrations with other writing tools or publishing platforms
Best For
Writing long-form fiction novels with AI-assisted outlining and draftingCreating and managing worldbuilding settings, character profiles, and plot arcsRewriting and refining chapter drafts for consistency, grammar, and styleReviewing chapters from a reader/editor perspective to identify pacing, character arc, and plot hole issuesImporting and referencing large volumes of existing fiction or reference material during writingOffline writing with local LLMs (via Ollama) for privacy-sensitive projects
FAQ
What is Vela?
Vela is an open-source, privacy-first AI-powered IDE for novel writing, designed for web fiction authors, indie writers, and creative professionals. It integrates LLM capabilities with a local RAG knowledge base and an automated novel writing pipeline.
How does Vela ensure privacy?
Vela is local-first: all data (novels, worldbuilding, knowledge base) is stored on your own computer using SQLite and a lightweight vector engine. LLM calls can be made locally via Ollama or through your own API keys (BYOK), so your content never leaves your machine unless you choose a cloud model.
What AI models does Vela support?
Vela supports OpenAI, DeepSeek, Gemini, Claude, Ollama (local offline), and GLM. It also allows intelligent task routing, e.g., using DeepSeek for outlines, Claude for polish, and a local model for privacy-sensitive reviews.
Can I use Vela without an internet connection?
Yes, if you use a local LLM (e.g., via Ollama) and store all data locally, Vela can operate fully offline. The RAG knowledge base and all writing data remain on your computer.
What platforms does Vela run on?
Vela is cross-platform: macOS (dmg), Windows (nsis), and Linux (AppImage).