Thread by @NousResearch

Nous Research presents Hermes 4, our latest line of hybrid reasoning models.

hermes4.nousresearch.com

Hermes 4 builds on our legacy of user-aligned models with expanded test-time compute capabilities.

Special attention was given to making the models creative and interesting to interact with, unencumbered by censorship, and neutrally aligned while maintaining state of the art level math, coding, and reasoning performance for open weight models.

You can try Hermes 4 in the new, revamped Nous Chat UI.

chat.nousresearch.com

Nous Chat has been reworked to include parallel interactions, completions mode, and a memory system, which is slowly being rolled out. We now provide a suite of open and closed models for this experience, from Hermes 4 to GPT-5.

For the first week, all Hermes 4 inference in Nous Chat is free of charge.

Alongside these models, Nous Research releases a technical report that details the entirety of its creation process.

arxiv.org/abs/2508.18255

The technical report includes a thorough set of evaluations of Hermes 4 and other popular LLMs, complete with the actual text-results of each test. We believe this report sets a new standard for transparency in benchmarking.

In pursuit of producing models that are open, steerable and capable of producing the full range of human expression, we created a new benchmark, RefusalBench, that tests a model’s willingness to be helpful in a variety of scenarios commonly disallowed by both closed and open models.

Hermes 4 achieves SOTA against all popular closed and open models in conforming to your values, without censorship.

Special thanks to our launch day partners - @chutes_ai, @nebiusai, and @luminal_ai - for serving these models and powering our new chat experience.

Check out their platforms for additional options for API inference.

And if you’re looking for support or a great AI community to join, check out our discord at discord.gg/NousResearch