LLMs and chatbots
The chat apps and assistants people actually open: ChatGPT, Claude, Gemini, Grok, Copilot, DeepSeek, and their peers. One row per product, not every model version.
Data last checked: · What changed
| Known for | Best foreditorial rating | Pros & cons | Official site |
|---|
How the ratings work
- Popularity is a 1–5 judgment (halves allowed) of how commonly a product shows up in recent traffic reports, app ranks, or usage leaderboards — relative to the other products on this page, not a computed formula.
- Captions name the source and a “checked Nov 2026” date. If there was no usable source, the caption says “Editorial estimate.”
- Best for is an editorial label of a typical use, not a test score.
- Cost is a five-step scale (Free / $ / $$ / $$$ / $$$$). See “How cost is rated” for the band definitions on this page. Captions cite a source or say the rating is an estimate.
- How popular a chatbot looks here is relative to the other assistants on this page, not a universal score. Website-visit shares under-count tools that live mainly inside an app or an office suite.
How cost is rated
- Cost is rated on the typical individual paid plan most people buy, per month, when we could find a published price this session. If the vendor marks a plan as popular or most popular, that is the plan we use. Otherwise we use the ordinary individual paid plan (not the cheapest limited tier, and not an ultra or teams tier). The caption names the plan. Annual “per month” equivalents are noted when the vendor publishes them. Higher ultra tiers do not set the band.
- Free — no individual paid plan is required for the core product.
- $ — under $15 / month.
- $$ — $15–29 / month.
- $$$ — $30–99 / month.
- $$$$ — $100+ / month, or quote-only with no public individual plan found this session.
- If the product is mainly a usage-based API, or no public individual price turned up, the caption says so and the band is an editorial estimate — never an invented dollar amount. Always confirm current pricing with the vendor.
Sources and limits
Primary web-traffic source: Similarweb estimates of visits to generative AI assistant websites in October 2026 (ChatGPT 55.7%, Gemini 24.2%, Claude 11.5%, DeepSeek 3.4%, Grok 2.2%, Microsoft Copilot 1.1%, Perplexity 0.8%). Those figures are website visits, not app sessions or people. OpenRouter weekly token rankings (mid-November 2026) are cited for API-side usage and do not measure consumer apps. Meta AI, Qwen, Mistral Le Chat, and Character.AI are not in that seven-site Similarweb set; their stars are editorial unless another source is named.
Independence
This site is ad-free and has no affiliate links. Official-site buttons go straight to the vendor. Stars are not paid placement.
About the icons
The first column is meant to look like the product’s app or site icon. Official vendor or App Store app icons are used when we could find one this session. If not, the tile is a generated initials icon in the same rounded-square style.