N
Hermes 4 405BOPEN
Hermes 4 is a large-scale reasoning model built on Meta-Llama-3.1-405B and released by Nous Research.
Prices update automatically — checked daily against provider list prices.
- $1.5 per 1M tokens, assuming 3 input for 1 output the convention used by Artificial Analysis
- cheaper than 9% of the 191 open-weight models tracked here
- 4 of 4 Nous Research models tracked here, from cheapest
Benchmarks & performance
Independent benchmark scores for Hermes 4 405B, measured by Artificial Analysis. Higher is better. Measurement mode: Reasoning.
Models at this level cost $0.22 per 1M tokens (median of 27) — this one costs $1.5.
Cheaper and better on this index: GLM 5.3 Flash · Gemini 3.7 Flash · Gemini 3.7 Flash (batch) and 127 more
About this model
Hermes 4 405B is an open-weight AI model by Nous Research. You can download and self-host it for free; the prices below are hosted-API list prices, tracked daily, for when you prefer convenience over self-hosting.
Frequently asked questions
What is Hermes 4 405B?
Hermes 4 405B is an AI language model from Nous Research. It is open-weight: you can download it and run it on your own hardware, for free. It scores 8.8 on the Artificial Analysis intelligence index.
Is Hermes 4 405B free?
The weights are free and open — you can self-host Hermes 4 405B and pay nothing per token. If you prefer a hosted API, list prices are $1 per million input tokens and $3 per million output tokens.
What is Hermes 4 405B good at?
Independent benchmarks from Artificial Analysis give it GPQA 72.7%, MMLU-Pro 82.9%, Humanity's Last Exam 10.9%, Long Context Reasoning 22.7%, LiveCodeBench 68.6%, SciCode 25.2%, AIME 2025 69.7%, IFBench 32.7%, τ²-Bench 22.2%, Terminal-Bench Hard 11.4%. It is particularly used for mathematical reasoning.
How fast is Hermes 4 405B?
It generates about 34 tokens per second, with a median 0.92s delay before the first token. Measured independently by Artificial Analysis.
Can I self-host Hermes 4 405B?
Yes. Hermes 4 405B has open weights, so you can download it and run it on your own GPU or server with tools like Ollama, vLLM or llama.cpp — with no per-token cost.