Agent power rankings · week 41

Which agents people are getting real work out of right now. Ranked by momentum from our own data: new use cases in 28 Sep 2026 – 4 Oct 2026 vs 21 Sep 2026 – 27 Sep 2026, their trend scores, and ‘worked’ votes. Snapshot taken 5 Oct 2026.

#MovementAgent and whyMomentum
1NEWClaude34 new use cases this week (32 the week before); top: Give agents documentation instead of memory plugins.161.1
2NEWCodex10 new use cases this week (14 the week before); top: Run a social publication engine from your agent.48.3
3NEWChatGPT Dots8 new use cases this week (0 the week before); top: Run your ChatGPT Dot as a project manager.45.2
4NEWHermes Agent4 new use cases this week (5 the week before); top: Give your Hermes agent Muse-style memory.19.2
5NEWMeta Muse3 new use cases this week (2 the week before); top: Give your Hermes agent Muse-style memory.14.5
6NEWPerplexity Computer3 new use cases this week (1 the week before); top: Get a weekly what-changed report with Perplexity Computer.13.1
7NEWGrok Bot3 new use cases this week (3 the week before); top: Make a meal-planning bot for your household.12.2
8NEWPi2 new use cases this week (1 the week before); top: Extend Pi with council, voice and TTS extensions.11.1
9NEWAmp2 new use cases this week (1 the week before); top: Build a big project mostly from your phone with Amp.10.0
10NEWManus1 new use case this week (0 the week before); top: Point Manus at a complex dataset.6.4
11NEWChatGPT agent1 new use case this week (0 the week before); top: Find every subscription in a year of receipts.6.4
12NEWOpenClaw1 new use case this week (0 the week before); top: Translate the WhatsApp groups your family relies on.4.9
13NEWGemini CLI1 new use case this week (1 the week before); top: Stop agents editing tests to make them pass.4.1
14NEWInstinct1 new use case this week (1 the week before); top: Text Instinct your weekly small-business chores.3.7
15NEWOpenCode1 new use case this week (1 the week before); top: Ticket in, PR out: four skills and a reviewer in CI.3.2

Momentum = 3 × new this week + 1.5 × max(0, this week − week before) + trend of this week's use cases ÷ 25 + 2 × ‘worked’ votes. Only use cases on agentsetups count, by their source date. This week was backfilled from our current data when rankings launched. No earlier week to compare with, so every agent is marked new. ‘Worked’ votes: D1 votes (I tried this → Worked).

ShareXLinkedInReddit

Archive