Hermes 4
Hermes 4
A large open model tuned to answer rather than to lecture you about the question.
Updated
The read
Nous fine-tunes open base models toward a steerable, low-refusal assistant, and Hermes 4 is the flagship of that work. It follows a system prompt further than any commercial model will, which matters for research, red-teaming, roleplay and any domain where the safety layers of the big labs get in the way of legitimate work. Serving something this large is expensive, and the responsibility for outputs moves to you.
Where it shines
- Follows a system prompt further than any hosted model
- Lowest refusal rate of anything in this index
- Long context and a strong tool-calling format
Worth knowing first
- Very large, so serving it is a real cost
- Fewer guardrails means the output risk is entirely yours
Index score
77.0
Provisional. This score is read from the product's public capability, not from the one-prompt rebuild the fully reviewed entries went through, so treat it as a placing rather than a verdict.
- Rank
- 37 of 53
- Builds
- LLMs
- Category
- Open weights
- Output
- Published weights, permissively licensed
- Pricing
- Free tier
- Best for
- Work the commercial models refuse for the wrong reasons
Scores are our own editorial judgement, weighted craft 55, speed 10, control 15 and value 20.
Also worth a look
Full index- 05Kimi K3Kimi K3ProvisionalThe largest open-weights model anyone has shipped, and it reads a million tokens without blinking.Open weightsFree tier87.5
- 06GLM-5.2GLM-5.2ProvisionalFrontier-class output, an MIT licence, and a million tokens of context. That combination is the story.Open weightsFree tier87.1
- 12Qwen 3.6Qwen 3.6ProvisionalThe open Qwen release most people should actually run, in sizes that fit real hardware.Open weightsFree tier83.3
