Skip to content
neuen
Menu
LiveKimi K3 and GLM-5.2 are now on neuen

One chat.
Every leading model.

Ask Qwen, Kimi, GLM and DeepSeek from a single thread. Switch models on any message and pay only for the replies you get.

Qwen 3.8 Flash

Coming soon · GPT, Claude, Gemini

8 models live today from

Alibaba QwenMoonshot AIZhipu AIDeepSeekGPT, Claude and Gemini coming soon

The app

Switch models mid-thread. Keep the whole conversation.

A real thread from neuen: Qwen 3.8 Max drafted the policy, then DeepSeek V4 Pro wrote the code with the full context. Each reply shows exactly what it cost.

neuen.ai/app
DeepSeek V4 Pro

I'm opening a small online shop in Hong Kong. Give me 4 short bullets my refund policy must cover.

Qwen 3.8 Max· 129 credits
  • When refunds are allowed — e.g., defective, damaged, misdescribed, or wrongly delivered items.
  • Time limit and condition — how many days customers have and what condition/packaging is required.
  • How to request and process — contact method, proof needed, and how long the refund takes.
  • Costs and exceptions — who pays return shipping and any non-refundable items or fees.

Write a Python function that checks whether an order is still inside a 14-day refund window.

Switched to DeepSeek V4 Pro
DeepSeek V4 Pro· 103 credits

Here's a Python function that checks whether an order is still within a 14‑day refund window from the order date.

from datetime import datetime, timedelta

def is_within_refund_window(order_date_str: str, date_format: str = "%Y-%m-%d") -> bool:
    """
    Check if an order is still within the 14‑day refund window.

    Args:
        order_date_str: Order date as a string.
        date_format: Expected format of the date string (default: YYYY-MM-DD).

    Returns:
        True if today's date <= order_date + 14 days, else False.
    """
    try:
        order_date = datetime.strptime(order_date_str, date_format).date()
    except ValueError:
        raise ValueError(f"Invalid date string: {order_date_str}. Expected format: {date_format}")

    today = datetime.now().date()
    deadline = order_date + timedelta(days=14)

    return today <= deadline

Example usage:

print(is_within_refund_window("2025-03-25"))  # True if today is 2025-04-08 or earlier

The function treats the order date itself as day 0 and includes the full 14th day, so the window is order_date ≤ today ≤ order_date + 14 days.

Message DeepSeek V4 Pro…↑

Replies captured from the live service on October 25, 2026.

What you get

Built for people who use more than one model.

Which is larger, 9.11 or 9.9? One sentence.

▾ Thought process · GLM-5.2

Analyze the Request: · Compare two decimal numbers: 9.11 and 9.9. · Constraint: Exactly one sentence. Compare the numbers: · 9.11: The first decimal digit is 1. · 9.9: The first decimal digit is 9. · 9 > 1, so 9.9 is larger than 9.11. Draft the sentence: · Draft 1: 9.9 is larger than 9.11 because 9 is greater than 1 in the tenths place. (Good, accurate, one sentence) · Draft 2: 9.9 is larger than 9.11 since nine-tenths is greater than one-tenth. (Also good) · Draft 3: 9.9 is larger than 9.11. (Simple, meets the constraint perfectly, but maybe a bit too simple? The prompt just asks "Which is larger", so answering directly is best.) Refine the sentence: · Let's add a brief explanation to be helpful while strictly maintaining the one-sentence constraint. · "9.9 is larger than 9.11 because 9 tenths is greater than 1 tenth." · Wait, is it clear enough? Yes. Final Polish: · "9.9 is larger than 9.11 because nine-tenths is greater than one-tenth." · Actually, simply stating "9.9 is larger than 9.11." fulfills the prompt without overcomplicating, but explaining why shows reasoning. Let's go with the one-sentence explanation. · "9.9 is larger than 9.11 because in the tenths place, 9 is greater than 1." Check Constraints: · Which is larger? Answered (9.9). · One sentence? Yes. Let's just give the most direct, clear answer. "9.9 is larger than 9.11 because 9 tenths is greater than 1 tenth."

9.9 is larger than 9.11 because nine-tenths is greater than one-tenth.

See how it thinks

Reasoning models show their working. GLM-5.2 on a classic trick question, collapsed by default.

1from datetime import datetime, timedelta
2
3def is_within_refund_window(order_date_str: str, date_format: str = "%Y-%m-%d") -> bool:
4 try:
5 order_date = datetime.strptime(order_date_str, date_format).date()
6 except ValueError:
7 raise ValueError(f"Invalid date string: {order_date_str}. Expected format: {date_format}")
8
9 today = datetime.now().date()
10 deadline = order_date + timedelta(days=14)
11
12 return today <= deadline

Code with context

DeepSeek V4 Pro picked up the thread and wrote working Python, formatted and ready to copy.

modelqwen-3.8-max
input tokens84
output tokens187
rate200,000 / 600,000 per 1M
charged129 credits

≈ US$0.0026

A receipt for every reply

Credits are charged on the tokens each reply actually used. 50,000 credits = US$1.

Pricing

Free every day. Pro when you need more.

Free

US$0forever

  • 5,000 credits every day
  • Every available model
  • Conversation history
Start free

Pro

US$20/ month

  • 1,000,000 credits every month
  • Plus 5,000 free credits every day
  • Every available model
  • Cancel any time from your account
Get Pro · US$20/month

Renews monthly until cancelled. Cancel any time.

Prefer no subscription? Credit packs from US$10. See full pricing

Your first 5,000 credits are on us. Every day.