An LLM (large language model) is, in everyday use, the AI you type or talk to in apps such as ChatGPT, Gemini or Claude: it answers questions, writes and edits text, reads your files and, more and more, carries out tasks on its own. The best LLM depends on what you need it for, which is why rankings disagree. The ABXK.AI Ranking brings the trustworthy ones together: independent blind comparisons, in which people vote on answers without knowing which LLM wrote them, and the tests of established tech publications. The LLMs that land in the top ten most consistently come first.
We have not tested these LLMs ourselves. How the ranking is built is explained in one sentence below the list.
The ABXK.AI Ranking
1. ChatGPT
The all-rounder that most reviews put first. Writing, web search, file analysis, voice chat, image generation and agent tasks come in one app, and many other tools connect to it. A safe default if one assistant should cover many jobs; our LLM database compares the models behind it.
Use ChatGPT to answer questions, write, create images, complete work, and code—all in one place. Get started for free or download the app.
2. Google Gemini
Strong when you work in Google’s world: on request it reads your Gmail, Docs and Drive files, and on Android it replaces the old Google Assistant. Reviewers point to its image generation and a deep research mode that writes long reports with sources.
Meet Gemini, Google’s AI assistant. Get help with writing, planning, brainstorming, and more. Experience the power of generative AI.
3. Grok
Grok can read posts on X directly, which helps with breaking news and trends, and it is built into the X app. Reviewers praise its reasoning and its trip planning. Treat claims from social media posts with care: popular is not the same as true.
Grok is an AI assistant built by SpaceXSI. Chat, create images, write code, and get real-time answers from the web and X.
4. Microsoft Copilot
The natural choice for teams on Microsoft 365: inside Word, Excel, Outlook and Teams it drafts, summarizes and analyzes with your company data, within the permissions you already have. The free consumer app covers chat, web search and images. Reviews rate it best for people who work with Microsoft.
Microsoft Copilot is AI built to help you get things done. Create, brainstorm, learn, and solve problems with AI chat that works across your day.
5. Claude
Reviewers single out Claude for writing quality and careful answers, and for coding: the same company makes the coding agent Claude Code. It handles long documents well and works with files, connected apps and the web. A good pick if most of your work is text.
Claude is a next generation AI assistant built by Anthropic and trained to be safe, accurate, and secure to help you do your best work.
6. Perplexity
Built for research rather than chat: every answer comes with numbered sources, and you can limit a search to academic papers or specific sites. Reviewers rate it best for verified search. Check the cited pages anyway: a source link does not guarantee that the summary is right.
Perplexity is a free AI-powered answer engine that provides accurate, trusted, and real-time answers to any question.
7. Meta AI
The assistant many people already have: it sits inside WhatsApp, Instagram and Facebook, so there is nothing to install. Reviewers see it as best for social media and quick everyday questions. Confidential business data, however, does not belong in chats inside Meta’s apps.
Use Meta AI to get things done. Create AI-generated images for free, edit photos, shop, find places to go, and get answers to any question. Start now.
8. DeepSeek
DeepSeek became known for strong reasoning models that anyone can download and run. The chat app itself is run by the company in China, and its privacy policy says that data is stored there; several European data protection authorities have acted against the app. For sensitive work, run the open model on your own or a European server instead.
DeepSeek is an AI research company focused on building world-leading general artificial intelligence. We develop and open-source frontier LLMs including DeepSeek-V4, DeepSeek-R1, and DeepSeek-Coder. Chat with DeepSeek AI or integrate via API.
9. Kimi
Moonshot AI releases its Kimi models with open weights, and the app handles long documents and agent tasks. The company is based in China, so check where your data is processed before you use it for work.
Try Kimi K3 to build playable multiplayer and 3D games, create consulting grade slides, and run parallel tasks with Swarm and Goal to get more work done.
10. Qwen Chat
Qwen Chat brings Alibaba’s latest Qwen models together in one app, including image and video generation. Many Qwen models are released with open weights and are among the most used open models for building your own apps. As with other providers based in China, check where your data is processed.
Qwen is an official platform from Qwen that empowers both everyday users and developers with unified access to Qwen’s series of open-source and proprietary models. It offers comprehensive functionality spanning chatbots, image and video understanding, image generation and editing, document processing, tool utilization, voice and video chat, and artifacts.
How the ABXK.AI Ranking works: we evaluate only well-known, reputable sources — independent blind comparisons and the tests of established publications — and order the tools by how often they appear in their top ten, with ties broken by the better average place. We have not tested the tools ourselves. As of October 18, 2026.
What the ranking shows
ChatGPT, Google Gemini and Grok are in the top ten of every source we evaluated. ChatGPT is ranked first most often by the tech publications, while Google’s model leads the blind comparison, where only the quality of the answers counts.
Microsoft Copilot, Claude and Perplexity follow closely. Their strengths are clearer than their overall placings: Copilot works inside Microsoft 365, Claude is praised for writing and coding, and Perplexity is built for research with sources. Meta AI and DeepSeek complete the top eight. Kimi and Qwen Chat enter the list through the blind comparison: their models from Moonshot AI and Alibaba score close to the leaders there, but the tech publications rarely review them yet.
LLM comparison: which one for what
| If you need | Look at | Why it stands out |
|---|---|---|
| One LLM for almost everything | ChatGPT | The broadest set of features and connections to other tools |
| Help inside Gmail, Docs and Android | Google Gemini | Connected to Google’s apps |
| Help inside Word, Excel, Outlook and Teams | Microsoft Copilot | Built into Microsoft 365 |
| Long texts, careful writing, coding | Claude | Reviewers praise its writing quality |
| Research with sources you can check | Perplexity | Numbered sources in every answer |
| News and trends from X | Grok | Live access to posts on X |
| Quick questions in WhatsApp or Instagram | Meta AI | Already inside the apps |
| Models you can also run yourself | DeepSeek, Kimi, Qwen Chat | Their models are released with open weights |
You can filter all LLMs on this list, and more tools, in the comparison table and put them side by side.
The app and the model inside
Each LLM on this list is an app with a language model inside, and the model is the engine. Most providers offer several models in the same app, from a fast one for everyday questions to a slower one that thinks longer, and they replace them every few months. So an LLM that disappoints you today may be better after the next model update, and two apps can run models from the same family. How the models themselves compare on benchmarks is shown in our LLM database, which is updated every week.
What to check before you share work data
LLMs are easy to try, which is exactly why work documents end up in them. Before you paste anything confidential, check three things:
- Training on your chats. Many consumer plans may use conversations to improve the models unless you switch this off. Business plans of the large providers usually exclude it by contract.
- Where the data is stored. DeepSeek, Kimi and Qwen Chat are run by companies based in China. DeepSeek’s privacy policy says that data is stored there, and several European data protection authorities have acted against its app.
- A contract for your company. If personal data is involved, your company needs a data processing agreement with the provider, and your team needs clear rules on what may go into an LLM.
More on safe use, from phishing to data leaks, is in AI Security.