A blank page is the slowest part of writing. This tool gets you to a structured first draft fast — give it a topic, the points you want covered, and a target length, and it returns a full Markdown blog post using your own OpenAI or Anthropic key, entirely in your browser.
How it works
Choose a provider and model, paste your API key, enter your topic, and optionally list outline points (one per line) and your target audience. Pick a word count. The tool builds a writing prompt that asks for a Markdown post with an H1 title, H2 sections, and a conclusion, sized to roughly your target length, and instructs the model to write specifically and avoid fabricating statistics. It sends one direct request to the provider and shows the draft with a live word count.
For Anthropic, the request includes the official direct-browser-access header so it works straight from the page.
What makes a usable first draft
The tool is designed to produce a draft worth editing rather than a polished final post. The difference matters: a draft that is structurally sound, covers the right points, and has clear section breaks cuts your writing time even if individual paragraphs need rewriting. The prompt deliberately asks for:
- A specific, descriptive H1 title rather than a generic one
- H2 sections that match your outline points, not the model’s default topics
- A concrete closing paragraph rather than a generic “in conclusion” filler
- A note about any statistic or figure that the model was uncertain about, so you know where to verify
The draft will always need human editing for voice, accuracy, and specific examples from your own experience. Treat it as a scaffold, not a publication.
Choosing your model and settings
For blog posts, the trade-off between model capability and cost is real. Smaller, cheaper models can produce competent first drafts for straightforward topics. Flagship models handle nuanced arguments, technical topics, and unusual angles better but cost more tokens per post. A reasonable approach is to use a smaller model for the first draft pass and a better model only for posts on complex or high-stakes topics.
Target word count affects cost roughly linearly: a 1,000-word post costs about twice as many tokens as a 500-word post. Set the target slightly above your intended final length since cutting is faster than padding.
Giving it more to work with
- Outline points steer structure — each line tends to become a section.
- Audience shapes vocabulary and examples — “café owners” reads very differently from “CFOs.”
- Word count sets the depth; longer posts cost more tokens on your key.
Tips
- Treat the output as a draft, not a finished post — edit for voice, cut filler, and verify every claim.
- Generate at a slightly higher word count than you need, then trim; cutting is faster than padding.
- Cheaper models produce serviceable drafts; reserve flagship models for posts where nuance really matters.
- Any statistics or figures the model produces should be checked against a real source before publishing — the prompt instructs it not to fabricate, but verify regardless.