Hermes 4.3
Hermes 4.3
A 36-billion-parameter model tuned to answer rather than to lecture you about the question.
Updated
The read
Hermes 4.3 is Nous Research's current release, built on ByteDance's open Seed-OSS 36B and trained to follow the user's instructions with as little moralising as possible. At 36 billion parameters it runs on far less hardware than the largest Hermes 4 it replaces, and it keeps a 512,000-token window. Apache-licensed and steerable through the system prompt, it is the open model people choose when commercial ones refuse for the wrong reasons.
Where it shines
- Follows the system prompt closely, with few reflexive refusals
- Apache 2.0 on a 36B model that fits a single large GPU
- 512,000 tokens of context
Worth knowing first
- A mid-size model, so the hardest reasoning is out of reach
- Light safety tuning puts the burden of guardrails on you
Index score
78.9
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
- 46 of 73
- Builds
- LLMs
- Category
- Open weights
- Output
- Apache-licensed weights, steerable by system prompt
- 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- 05GLM-5.3 FlashGLM-5.3 FlashProvisionalA fast, MIT-licensed open model that human voters rank beside the big flagships.Open weightsFree tier89.6
- 06Xiaomi MiMo V2.6 ProMiMo V2.6 ProProvisionalThe strongest large open-weights model on the independent index, under MIT, for under a dollar out.Open weightsFree tier89.4
- 07DeepSeek V4.1 FlashDeepSeek V4.1 FlashProvisionalFrontier-adjacent answers for thirty cents per million input tokens, with the weights under MIT.Open weightsFree tier88.3
