Subscribe to our newsletter
One short email when we publish. No spam, unsubscribe anytime.

Qwen4-Max Is Free on the Nous Portal for Five Days, and Hermes Agent Can Use It
Alibaba's newest large AI model, Qwen4-Max, can be used free of charge on the Nous Portal for five days, Nous Research announced on the 18th of November 2026. The Nous Portal is one of the model services Hermes Agent connects to, according to the Hermes Agent documentation, so a Hermes Agent user can point the agent at Qwen4-Max during the offer and try it on real work.
Nous Research made the announcement in a post on X on the 18th of November, saying the offer runs "in partnership with" the Alibaba Qwen team and the company Inferact. The post says five days and gives no exact end date. By the morning of the 20th of November it had been viewed more than 517,000 times, according to the post's public counters.
What Qwen4-Max is, in plain words
A model is the "brain" an AI agent thinks with. The agent carries out the work, and the model decides what to say and do next. Changing the model is like giving the same assistant a different way of thinking.
Alibaba presented Qwen4-Max on the 16th of November 2026 in a press release, calling it the most capable Qwen model it has built so far. The release says the model can read up to one million tokens at once. A token is a small piece of a word, so the model can keep a very long document or conversation in view at the same time.
The Nous Research post adds two points: the model has native vision, which means it can read screenshots and photos along with text, and it is built for long-horizon work, the long multi-step jobs an agent runs on its own.
How Hermes Agent users switch to it
According to the Hermes Agent documentation published by Nous Research, an agent connects to the Nous Portal in three moves:
- Choose Nous Portal as the model provider in Hermes Agent.
- Sign in to the Nous Portal in the browser window that opens.
- Pick a model from the portal's list. On the 20th of November the Nous Portal models page showed "Qwen: Qwen4 Max" in its section of free models, marked FREE.
The same documentation describes a model picker inside a chat session, so the switch can also happen in the middle of a conversation.
What the paid version costs
The Nous Portal models page also lists a paid version of Qwen4-Max, at the following prices read on the 20th of November 2026. Euro amounts use the European Central Bank reference rate of the 19th of November, 1 euro to 1.0992 dollars.
| Qwen4-Max on the Nous Portal | Price per million tokens |
|---|---|
| Text the model reads (input) | $1.42 (€1.29) |
| Text the model writes (output) | $4.27 (€3.88) |
The same page shows a list price of $1.50 (€1.36) for input and $4.50 (€4.09) for output, with the Nous Portal price set about 5 percent below it.
Why the five days matter
A free window is the cheapest moment to see how a new model handles the jobs an agent already does: a morning summary, a weekly report, a long research task. Those jobs run on a schedule, and a scheduled job only runs while the agent is switched on. A Hermes Agent on a server that runs day and night keeps running its scheduled jobs on the new model through the night and over the weekend, while its owner's computer is closed.
The announcement is in the Nous Research post on X, and the prices are on the Nous Portal models page.
Source: x.com
Helpful documentation
Want your own AI agent running without the setup? We host it on a dedicated EU server. Free for 7 days, no credit card.
Start your free trialGet notified when we publish new articles
No spam, unsubscribe anytime.
By subscribing, you agree to our Privacy Policy.
