--- title: Hermes 2 Pro Llama-3 8B — Function Calling LLM | ModelsLab description: Access NousResearch: Hermes 2 Pro - Llama-3 8B for precise function calling and JSON outputs. Try this 8B LLM API now. url: https://modelslab-frontend-v2-927501783998.us-east4.run.app/nousresearch-hermes-2-pro-llama-3-8b canonical: https://modelslab-frontend-v2-927501783998.us-east4.run.app/nousresearch-hermes-2-pro-llama-3-8b type: website component: Seo/ModelPage generated_at: 2026-05-13T10:37:20.577987Z --- Available now on ModelsLab · Language Model NousResearch: Hermes 2 Pro - Llama-3 8B Function Calling Perfected --- [Try NousResearch: Hermes 2 Pro - Llama-3 8B](/models/open_router/nousresearch-hermes-2-pro-llama-3-8b) [API Documentation](https://docs.modelslab.com) Master Tools and JSON --- Core Upgrade ### Retrain on OpenHermes 2.5 Uses cleaned OpenHermes 2.5 dataset plus in-house function calling data for reliable outputs. 90% Accuracy ### Excels at Function Calls Scores 90% on Fireworks.AI eval; supports multi-turn with ChatML and special tokens like . JSON Mastery ### Structured Outputs Achieves 84% on JSON eval; enables agentic parsing during streaming with single-token tags. Examples See what NousResearch: Hermes 2 Pro - Llama-3 8B can create --- Copy any prompt below and try it yourself in the [playground](/models/open_router/nousresearch-hermes-2-pro-llama-3-8b). JSON Weather Query “<|im\_start|>system You are a helpful assistant with access to tools. Use JSON for structured responses.<|im\_end|> <|im\_start|>user Get current weather in Tokyo using tool.<|im\_end|> <|im\_start|>assistant get\_weather{"city":"Tokyo"}<|im\_end|>” Multi-Turn Calc “<|im\_start|>system Handle math via calculator tool. Respond in JSON.<|im\_end|> <|im\_start|>user Calculate 25 \* 4 + 17.<|im\_end|> <|im\_start|>assistant {"expr":"25\*4+17"}<|im\_end|>” Code Gen Task “<|im\_start|>system Generate Python code. Output as JSON with reasoning.<|im\_end|> <|im\_start|>user Write function to sort list ascending.<|im\_end|> <|im\_start|>assistant {"reasoning":"Use sorted() builtin.", "code":"def sort\_list(lst): return sorted(lst)"}<|im\_end|>” Agent Workflow “<|im\_start|>system Parse tools for agent tasks. Use after calls.<|im\_end|> <|im\_start|>user Book flight to NYC on tool.<|im\_end|> <|im\_start|>assistant search\_flights{"dest":"NYC"}<|im\_end|>” For Developers A few lines of code. JSON Tools. ChatML Format. --- ModelsLab handles the infrastructure: fast inference, auto-scaling, and a developer-friendly API. No GPU management needed. - **Serverless:** scales to zero, scales to millions - **Pay per token,** no minimums - **Python and JavaScript SDKs,** plus REST API [API Documentation ](https://docs.modelslab.com) PythonJavaScriptcURL Copy ```

import requests

response = requests.post(
    "https://modelslab.com/api/v7/llm/chat/completions",
    json={
  "key": "YOUR_API_KEY",
  "prompt": "",
  "model_id": ""
}
)
print(response.json())

``` FAQ Common questions about NousResearch: Hermes 2 Pro - Llama-3 8B --- [Read the docs ](https://docs.modelslab.com) ### What is NousResearch: Hermes 2 Pro - Llama-3 8B? Hermes 2 Pro is an 8B Llama-3 fine-tune by NousResearch. It upgrades Hermes 2 with function calling and JSON datasets. Excels in general tasks plus 90% function calling accuracy. ### How does nousresearch hermes 2 pro llama 3 8b handle function calling? Uses special ChatML with , tokens as single tokens for streaming. Supports multi-turn via system prompts. Scores 90% on Fireworks.AI eval. ### What is NousResearch: Hermes 2 Pro - Llama-3 8B API context length? Supports 8192 input tokens and 8192 max output. Uses Llama3 tokenizer in ChatML format. Ideal for structured agent workflows. ### Is NousResearch: Hermes 2 Pro - Llama-3 8B model good for JSON outputs? Yes, scores 84% on structured JSON eval. Employs in-house dataset for reliable parsing. Adds tokens like for agents. ### Where to find NousResearch: Hermes 2 Pro - Llama-3 8B alternative? Hosted on Hugging Face and OpenRouter. GGUF quant versions available for llama.cpp. Check Novita or similar for API access. ### What prompting for nousresearch hermes 2 pro llama 3 8b api? Use ChatML with roles and special tokens. System prompts guide behavior. More structured than Alpaca for multi-turn dialogue. Ready to create? --- Start generating with NousResearch: Hermes 2 Pro - Llama-3 8B on ModelsLab. [Try NousResearch: Hermes 2 Pro - Llama-3 8B](/models/open_router/nousresearch-hermes-2-pro-llama-3-8b) [API Documentation](https://docs.modelslab.com) --- *This markdown version is optimized for AI agents and LLMs.* **Links:** - [Website](https://modelslab.com) - [API Documentation](https://docs.modelslab.com) - [Blog](https://modelslab.com/blog) --- *Generated by ModelsLab - 2026-05-13*