Claude#2
Anthropic's assistant, weighted towards longer context, nuanced writing and careful reasoning, with document analysis and coding among its stated strengths. Freemium, and the closest peer here on score.
ChatGPTChatbots & AssistantsCurrently ranked #1Updated September 2026
ChatGPT takes the top slot in Chatbots & Assistants with a composite score of 60.0, on a freemium plan. It covers writing, research, coding help and everyday tasks, with plugins, custom GPTs and voice. The nine below hold the rest of the category and differ less in what they answer than in where they live and what they answer from: your inbox, your messaging apps, the live web, or open weights you could run yourself. Scores run 57.0 down to 33.0.
Some links on this page are affiliate links, marked nofollow sponsored in the page source. They may earn us a commission at no cost to you, and they never affect the ranking — every position comes from the published composite score.
Anthropic's assistant, weighted towards longer context, nuanced writing and careful reasoning, with document analysis and coding among its stated strengths. Freemium, and the closest peer here on score.
Answers with citations pulled from the live web, which makes it research-first rather than chat-first. Pick it when you have to show where an answer came from.
Sits inside Gmail, Docs and Workspace and handles text, image and voice in one place. The obvious candidate if your working day already happens in Google.
| TOOL | RANK | SCORE | USER RATING | PRICING |
|---|---|---|---|---|
| ChatGPTCOMPARED TOOL | #1 | 60.0 | Not rated yet | FREEMIUM |
| Claude | #2 | 57.0 | Not rated yet | FREEMIUM |
| Perplexity | #3 | 54.0 | Not rated yet | FREEMIUM |
| Google Gemini | #4 | 51.0 | Not rated yet | FREEMIUM |
| Poe | #5 | 48.0 | Not rated yet | FREEMIUM |
| Microsoft Copilot | #6 | 45.0 | Not rated yet | FREEMIUM |
| Grok | #7 | 42.0 | Not rated yet | FREEMIUM |
| Meta AI | #8 | 39.0 | Not rated yet | FREE |
| DeepSeek | #9 | 36.0 | Not rated yet | FREEMIUM |
| Mistral Le Chat | #10 | 33.0 | Not rated yet | FREEMIUM |
Anthropic's assistant, weighted towards longer context, nuanced writing and careful reasoning, with document analysis and coding among its stated strengths. Freemium, and the closest peer here on score.
Read the Claude listing →Answers with citations pulled from the live web, which makes it research-first rather than chat-first. Pick it when you have to show where an answer came from.
Read the Perplexity listing →Sits inside Gmail, Docs and Workspace and handles text, image and voice in one place. The obvious candidate if your working day already happens in Google.
Read the Google Gemini listing →One subscription that reaches ChatGPT, Claude, Gemini and dozens of other models. Useful when you would rather compare answers than commit to a single house style.
Read the Poe listing →Built into Windows, Office and Edge, so the help arrives in the software your company already licenses rather than in another tab you have to open.
Read the Microsoft Copilot listing →xAI's assistant, tied closely to X and known for reaching real-time information. Relevant when the question is about what is happening rather than what is documented.
Read the Grok listing →Embedded across WhatsApp, Instagram and Messenger, with image generation alongside conversation. The only entry in this ten that is free outright.
Read the Meta AI listing →A free chat assistant on open-weight reasoning models, with a reputation built on cost-efficient performance rather than on breadth of features.
Read the DeepSeek listing →Mistral's assistant on its own open-weight models, emphasising speed and European hosting. That last point is the deciding factor for some buyers and irrelevant to others.
Read the Mistral Le Chat listing →Most people switch for one of three reasons. Research with citations points at Perplexity. Wanting the assistant inside tools you already have points at Gemini for Workspace, Microsoft Copilot for Windows and Office, or Meta AI for WhatsApp and Instagram. Wanting several models without several subscriptions points at Poe. Nine of these ten are free to try, Meta AI entirely so, which makes this an unusually cheap category to test properly.
Claude ranks highest among the alternatives here at 57.0, against ChatGPT's 60.0. For sourced research rather than conversation, Perplexity at 54.0 is the different answer worth considering.
Meta AI is free. Claude, Perplexity, Gemini, Poe, Microsoft Copilot, Grok, DeepSeek and Mistral Le Chat are all freemium, so every tool in this top ten can be tried without paying.
Perplexity answers with citations pulled from the live web, which is the reason it sits in this category as a research-first option.
By the composite score published on the How we rank page. Every listing in the directory is scored on the same formula, and each category holds ten slots.