Doc to Blog Studio

Status: Live.

1. Your muse

   Rebuild ofGenerating technical blogs from your docs with Llama and RAG (opens in a new tab)Clean

   Turn my product docs into a technical blog post with a header image. This rebuilds Meta's video "Generating technical blogs from your docs with Llama and RAG" (https://dev.meta.ai/llama/resources/videos/generate-technical-blogs-from-your-docs-with-Llama-and-RAG) on current Muse models. It demonstrates: Drafting blog posts from existing documentation with retrieval. Plan: Upload or paste docs and get a draft post with a header image. Most doc sets fit in Spark 1.3's 1M-token context, so there is no retrieval pipeline to build.

2. Show thinking

3. Checks passed

4. 

   The plan: Doc to Blog Studio

   Turn product docs into a technical blog draft with header image

   1. Paste docs or upload PDF files
   2. Pick blog topic, audience and tone
   3. Generate draft post from docs
   4. Generate header image for the draft

   See the full plan

   Who it's for

   Developers and technical writers turning docs into posts

   How it works

   1. Paste docs or upload PDF files
   2. Pick blog topic, audience and tone
   3. Generate draft post from docs
   4. Generate header image for the draft
   5. Edit text and download the result

   Screens

   • Docs Input. Paste text and upload PDFs and set topic and tone
   • Draft Editor. Read and edit generated post and header image
   • Export. Copy text and download image and markdown file

   What goes in

   • Text: Pasted product docs plus topic, audience and tone
   • PDF (optional): Optional uploaded doc files under the size limit

   What comes out

   • Text: Markdown blog draft with title, headings and code snippets
   • Image: Landscape header image for the post

   Models

   Model	Used for	Why
   Muse Spark 1.3	Draft blog post from docs	Reads pasted text and PDFs in large context and writes structured draft with headings and code
   Muse Image	Create header image	Generates original header art from post topic and style prompt

   Limits

   • Rebuild fit: clean. Same drafting demo as video but uses Spark 1.3 large context instead of a separate retrieval pipeline and uses current image model.
   • Large doc sets over about 900k tokens are truncated to fit context.
   • Drafts may invent details. User must check facts against source docs.
   • Header images are original art and do not copy video footage or logos.

   From your inspo

   Rebuild of Meta video Generating technical blogs from your docs with Llama and RAG. Keeps uploading docs to draft a technical post with header image using current models.

   Takes

   • Drafting blog posts from existing documentation
   • Creating a header image for the post

   Leaves out

   • Video footage and script
   • Meta and Llama names and logos
   • Exact look of any real product shown

   Cost per run: about $0.04 (est.). About 15k input tokens at $1.25 per M is $0.019, about 1800 output tokens at $4.25 per M is $0.008, plus one image at $0.01. Total about $0.04.

6. Done. npm run verify exits 0: typecheck, build, start, console, smoke (4 passed), and manifest all green.

   It's live.

   musemuse-doc-to-blog-studio.vercel.app (opens in a new tab)

   Product pageNew muse

   5 m 32 s · Input 2.2M · Output 42.1k · $0.76 est.