Hermes Models by Nous Research
About
The Hermes family of Large Language Models (LLMs), crafted by Nous Research, is based on Meta's Llama 3.1 framework. These models are noted for their high level of steerability and customization, setting them apart from many proprietary counterparts. Available in model sizes ranging from 8 billion to 405 billion parameters, the Hermes LLMs possess advanced agentic capabilities, impressive role-playing skills, and enhanced reasoning and creativity. The latest iteration, Hermes 3, stands out for its unique behavior, sometimes showing "existential crises" during certain interactions. This feature highlights the intricate nature of modern AI models as they scale. As open-source models available on Hugging Face, Hermes LLMs are accessible for community-driven adjustments, promoting powerful AI tools aligned with user needs without corporate limitations.
Current Variants
Use-when guidance is derived from seed capabilities, context, release, and replacement fields.
Use when the workload needs 13B parameters and structured outputs.
Use when the workload needs 7B parameters and structured outputs.
| Model | Use when | Released | Signals | Status |
|---|---|---|---|---|
| Nous Hermes Llama 2 13B | Use when the workload needs 13B parameters and structured outputs. | 2023-12 | 13B parametersstructured outputs | Current |
| Nous Hermes Llama 2 70B | Use when the workload needs 70B parameters. | 2023-12 | 70B parameters | Current |
| Nous Hermes Llama 2 7B | Use when the workload needs 7B parameters and structured outputs. | 2023-12 | 7B parametersstructured outputs | Current |
| Nous Hermes 13B | Use when the workload needs 13B parameters. | 2023-12 | 13B parameters | Current |
Release Timeline
1 release groupSpecifications(4 models)
| Model | Released | Parameters | Structured Outputs |
|---|---|---|---|
| Nous Hermes Llama 2 13B | 2023-12 | 13B | Yes |
| Nous Hermes Llama 2 70B | 2023-12 | 70B | No |
| Nous Hermes Llama 2 7B | 2023-12 | 7B | Yes |
| Nous Hermes 13B | 2023-12 | 13B | No |
Available From(4 providers)
Pricing
| Model | Provider | Input / 1M | Output / 1M | Type |
|---|---|---|---|---|
| Nous Hermes Llama 2 13B | Replicate API | $0.1 | $0.5 | Serverless |
| Nous Hermes 13B | Lepton AI API | $0.13 | $0.13 | Serverless |
| Nous Hermes Llama 2 7B | Together AI | $0.2 | $0.2 | Serverless |
| Nous Hermes Llama 2 13B | Fireworks AI | $0.2 | $0.2 | Provisioned |
| Nous Hermes Llama 2 7B | Fireworks AI | $0.2 | $0.2 | Provisioned |
| Nous Hermes Llama 2 13B | Together AI | $0.3 | $0.3 | Serverless |
| Nous Hermes Llama 2 70B | Fireworks AI | $0.9 | $0.9 | Provisioned |
Frequently Asked Questions
- What is Hermes used for?
- Hermes is used for structured outputs, coding, and agent workflows. The family description and listed model capabilities point to those workloads as the best fit.
- How does Hermes compare to MOSS-Audio?
- Hermes by Nous Research is strongest where you need structured outputs, while MOSS-Audio by MOSI Intelligence is the closest related family to check for multimodal. Hermes has 4 listed variants, so compare the specs and pricing tables before choosing a production model.
- Which Hermes model should I use?
- For the lowest listed input price, start with Nous Hermes Llama 2 13B through Replicate API at $0.1/1M input tokens. For the most capable/latest local choice, evaluate Nous Hermes Llama 2 13B with structured outputs.
