No single AI does everything. We sorted 686 combos that split the work between a thinking side (the brain) and a making side (executor tools) into 11 fields. The combos use 85 tools in total; the tools most often used as the brain are ChatGPT (83×), Claude (83×), Gemini (81×), Perplexity (77×), DeepSeek (76×).
153 of the 686 combos have an impact score of 10. Here are the first 6; see them all in the field groups.
Hook-driven product comparison short-form video for Reels and TikTok
★★★★★ 10/10Claude designs, at length, the scripts for each target audience, the scene lists, and the prohibited expressions. Only the camera and motion prompts for each scene are passed to Higgsfield, so image and video generation costs are kept separate. Make or n8n queues only approved scripts, and views and completion rates feed into the next hook selection.
Pricing Claude free/Pro·Max + API pay-as-you-go, Higgsfield free trial/credits·subscription
Official pricing page collection and parsing pipeline
★★★★★ 10/10Claude explains changes in each site's HTML, JSON-LD, and API responses, and Cursor implements a testable parser. The collector separates robots.txt, terms of service, request rate, and failure retries into policies, and excludes non-public areas. Change detection results are compared with previous prices and sent to a review queue, leading to human approval instead of automatic publishing.
Pricing Claude free/Pro·Max + API pay-as-you-go, Cursor free/Hobby·Pro·Business
Automation from price change detection to content drafts
★★★★★ 10/10Make receives change signals from approved official URLs and asks ChatGPT for a summary of the differences. New prices are saved in Airtable or a DB with a review status, along with the source snapshot, collection time, and confidence level. Publishing, sending emails, and updating ad copy run only after human approval, preventing the spread of incorrect changes.
Pricing ChatGPT free/subscription + API pay-as-you-go, Make free/subscription based on run count·workflows
On-premises-friendly pricing data review workflow
★★★★★ 10/10n8n connects crawl results, change logs, and approvers' decisions into a single workflow. Claude classifies whether a detected DOM change is an actual price change or just a design tweak. In a self-hosted environment, secret keys, access permissions, and audit logs are separated to meet enterprise data handling requirements.
Pricing Claude free/Pro·Max + API pay-as-you-go, n8n free/subscription based on run count·workflows
Source-tracking pricing update agent
★★★★★ 10/10Agents split collection, field extraction, change comparison, and review requests by role. n8n orchestrates the input, output, retries, and approvals of each step and reads API keys from a secure credential store. Automation success is managed not by the volume published but by source completeness, approval rate, and error correction time.
Pricing Agents SDK open source, connected LLM API and n8n usage billed separately
From change detection to publishing after approval
★★★★★ 10/10ChatGPT (OpenAI) links changes in official URLs to DB records. In Make, difference summaries, review requests, and publishing tasks are orchestrated in order. External publishing and sending are blocked until a person approves.
Pricing Brain: API pay-as-you-go + subscription (ChatGPT) / Tool: freemium (free + paid subscription)
Source: AI Galaxy Map combo data (ai_combos.json: brain, executor tools, workflow, pricing summary and impact for each combo) · Data compiled 2026-10-15 · pricing summaries are compiled from each tool's official pricing and can change