Try Qwen-Max for free and compare it with another AI model on your task in 7 minutes
No installation or card needed - two browser tabs and one prompt. In the end, you'll know which model handled your task better.
- 1.Open chat.qwenlm.ai - Alibaba's official free platform for Qwen. Click «Sign In» (you can log in via Google or GitHub). In the model list on the left, select Qwen-Max.
The site is available worldwide without VPN. Confirmation emails sometimes land in spam.
Qwen Chat ↗ - 2.Open chat.lmsys.org in a new tab - a free platform for testing AI models without registration. Select the «Direct Chat» tab at the top. From the dropdown, choose a competitor model to compare (for example, Claude-3.5-Sonnet or GPT-4o).
If the model you need isn't in the list, it's temporarily unavailable. Try again later or pick the closest alternative.
LMSYS Chatbot Arena ↗ - 3.Prepare one practical task - for example, text to edit, code with a bug, or a business email. Paste this prompt into Qwen-Max and hit send:
Don't use confidential data - both services process text on their servers.
Qwen Chat ↗PromptYou are an expert editor. Read the following text and rewrite it to be clear, concise, and professional. Fix grammar and logic errors. Keep the original meaning. Output only the revised text, no explanations. Text: [insert your text here, in English or Russian]
What this prompt doesThe prompt asks the model to edit the text: fix errors, make the style professional, and preserve the meaning. Replace [insert your text here...] with your actual text. - 4.Copy Qwen-Max's response into a notepad or Google Docs and label it «Qwen». Then paste the same prompt into LMSYS in the competitor model's tab and send it.
Use the exact same prompt text - otherwise the comparison won't be fair.
LMSYS Chatbot Arena ↗PromptYou are an expert editor. Read the following text and rewrite it to be clear, concise, and professional. Fix grammar and logic errors. Keep the original meaning. Output only the revised text, no explanations. Text: [insert your text here, in English or Russian]
What this prompt doesSame prompt as in step 3 - for a fair comparison. - 5.Compare both responses on three criteria: 1) accuracy - is the meaning preserved; 2) quality - how much better is the style vs. the original; 3) clarity - which text would you actually send. Note the winner for each point.
If the task is creative, repeat the test on 1-2 more different texts for a more reliable conclusion.
- 6.For the second test, repeat steps 3-4 with a different type of task. If the first test was about text, now give a coding, data analysis, or logic task. This will show whether the win was one-sided.
Models are often strong in one area and weak in another. Two tests give a fairer picture.
Qwen Chat ↗PromptYou are a senior Python developer. Find and fix the bug in the following code. Explain the bug in one sentence, then output the corrected code. Code: [insert your code here]
What this prompt doesThe prompt asks to find and fix a bug in the code and explain the issue in one sentence. Replace [insert your code here] with an actual code snippet containing a bug.
- freeQwen Chat (chat.qwenlm.ai)Completely free at the moment, no card verification required; the Qwen-Max model is available with no limits on the number of requests.
- freeLMSYS Chatbot Arena (chat.lmsys.org)Free, no registration needed; during peak hours there may be queues or some models may be temporarily unavailable.
Why today
AI services change fast - interfaces and free limits may differ from what's described.