# Kinsy > Kinsy is an AI presentation builder: it plans a narrative from your > notes and sources, runs live research with citations, fact-checks the > slides against your material, generates on-brand visuals, and exports > a native, editable PowerPoint (.pptx) — not screenshots. Pricing is credit-based: every plan grants monthly credits, actions (deck builds, edits, polish passes, AI images) consume them, and one-time top-up packs never expire. ## Pages - [Pricing](https://kinsy.ai/pricing): plans, credit menu, and top-up packs. - [Blog](https://kinsy.ai/blog/): comparisons and alternatives across the AI deck makers (Gamma, Beautiful.ai, Canva, Tome, Pitch), guides on turning documents into decks, and deep dives on native PPTX export fidelity. - [How to use Kinsy](https://kinsy.ai/blog/how-to-use-kinsy/): the full walkthrough from brief to exported PowerPoint, with credit costs per step. - [How to make a presentation with AI](https://kinsy.ai/blog/how-to-make-a-presentation-with-ai/): the complete workflow — five input routes, tool types, and the export test. - [Editable PPTX export compared](https://kinsy.ai/blog/ai-presentation-maker-editable-pptx/): which AI presentation makers export truly editable PowerPoint files, and the five-minute audit to verify any of them. - [Contact](https://kinsy.ai/contact): contact form; support email support@kinsy.ai. - [Terms](https://kinsy.ai/terms) · [Privacy](https://kinsy.ai/privacy): decks are private by default; user content is never used to train AI models. - [Full content](https://kinsy.ai/llms-full.txt): the complete text of every marketing page and blog post in one plain-text file. ## Key facts - Decks are planned by a frontier reasoning model, slide by slide. - Optional live web research attaches dated, numbered citations. - A fact-check pass verifies slide claims against the source material. - Exports are native PowerPoint/PDF/PNG; decks stay editable. - Free tier includes 100 credits (about one full deck). ## Developers - [API overview](https://kinsy.ai/docs): create finished, editable .pptx decks over a small JSON API at https://kinsy.ai/v1; keys spend the same credits as the web app, no separate API plan; limits and pricing in plain text. - [Quickstart](https://kinsy.ai/docs/quickstart): brief in, .pptx out in four curl calls; a finished deck takes one to two minutes. - [Command line](https://kinsy.ai/docs/cli): the kinsy CLI (npm `kinsy-cli`, Node 20+): `npx kinsy-cli login` once, then `npx kinsy-cli create "…" --file report.pdf --wait --out deck.pptx`; every flag and exit code. - [Use Kinsy from your agent](https://kinsy.ai/docs/agents): one Agent Skill + CLI, installable with `npx skills add kinsy-ai/kinsy-agents`, `claude plugin install kinsy@kinsy-ai` or Codex's marketplace; source at https://github.com/kinsy-ai/kinsy-agents (MIT). - [MCP server](https://kinsy.ai/docs/mcp): remote MCP server at https://mcp.kinsy.ai/mcp (Streamable HTTP, API-key bearer); seven tools: create_deck, wait_for_deck, get_deck, export_deck, send_to_google_slides, edit_slide, list_themes; setup for Claude Code, Codex, Cursor and VS Code. - [Authentication](https://kinsy.ai/docs/authentication): bearer API keys from Settings, the X-Kinsy-Client and Idempotency-Key headers, rate limits, signed download links. - [API reference](https://kinsy.ai/docs/api): every /v1 call, field by field; machine-readable at https://kinsy.ai/v1/openapi.json. - [PowerPoint add-in](https://kinsy.ai/docs/powerpoint): build a deck from a brief and documents inside PowerPoint (web, Windows, Mac) and insert its slides after the current slide; install from https://kinsy.ai/addin/manifest.xml. - [Google Slides](https://kinsy.ai/docs/google-slides): connect Google Drive once in Settings, then POST /v1/decks/{id}/deliver {"to": "google_slides"} (or the MCP send_to_google_slides tool, or `kinsy get --slides`) returns a Google Slides link in the user's own Drive. - [Webhooks and live progress](https://kinsy.ai/docs/webhooks): a signed callback when a deck or a whole-deck edit finishes, how to verify it, and a server-sent event stream of a build's progress. - [Errors](https://kinsy.ai/docs/errors): one error envelope with a stable code, a human message, and a checkout link on 402.