Description
Raptor Write is a bring-your-own-key writing application built around a plain document editor rather than a chat window. Instead of running on one built-in model, it connects directly to a writer's own API keys for OpenRouter, OpenAI, Anthropic, Google, MistralAI, DeepSeek, Perplexity, XAI, and Cohere, so the writer chooses which provider and model generates or continues their text. Through OpenRouter specifically, the settings panel also groups models into lists such as a top-10 set and a free-models list, letting a writer switch quickly between options without leaving the editor.
Read More
Inside the editor, writing is organized into projects and documents, with saved custom prompts that can be triggered from the keyboard to continue or revise selected text. Model settings are exposed directly, including temperature, output length (from a few hundred tokens up to over 100,000), and, on supporting providers, reasoning effort or extended-thinking token budgets. An Autocontinue feature lets generation keep running without repeated manual prompts, and an "Interns" feature is available alongside prompts and projects in the workspace.
Raptor Write also builds in a Pomodoro sprint timer with configurable settings, aimed at writers who work in timed focus blocks, plus hotkey-driven controls for running prompts, confirming AI output, and starting a new document. Finished work can be exported or saved as Markdown, plain text, or Word format, and prompts themselves can be exported and imported so a writer's setup can move between sessions or devices. A Notion API integration is listed as coming soon but isn't active yet.
Because the app requires the writer's own API keys for whichever provider they choose, there's no built-in free generation, usage costs run directly through each provider's own billing, and no pricing is charged by Raptor Write itself for the editor. The tool is built for writers who already have their own LLM API access and want a single, model-agnostic writing environment with structured prompts, timers, and multi-format export, rather than a hosted, all-in-one AI writing subscription.
- CategoryAI Writing Assistants
- Pricing ModelFree
- DeploymentCloud
- PlatformsWeb
Raptor Write's Core Features
Bring-your-own-key access to OpenRouter, OpenAI, Anthropic, Google, MistralAI, DeepSeek, Perplexity, XAI, and Cohere
Document editor organized into projects and saved documents
Custom, saved prompts triggered by keyboard hotkeys
Per-provider model settings: temperature, output length, reasoning effort, thinking budget
Autocontinue for uninterrupted generation without repeated prompting
Built-in Pomodoro sprint timer with configurable settings
Keyboard shortcuts for running prompts, confirming output, and creating documents
Export and import of documents (Markdown, plain text, Word) and of saved prompts
How to use Raptor Write?
Add your API key: Enter your key for OpenRouter or another supported provider.
Pick a provider and model: Choose which AI model will generate or revise text.
Start a project: Create documents and save custom prompts for your workflow.
Write with AI assistance: Use hotkeys to continue or revise selected text.
Time your sessions: Run the built-in Pomodoro sprint timer while you write.
Export your work: Save finished documents as Markdown, plain text, or Word.
Raptor Write's Use Cases
- Long-form drafting
- Focused writing sprints
- Prompt-driven revision
- Multi-model comparison
