Hermes 4 – Llama-3.1 405B (Reasoning)
Analysis Summary
Hermes 4 405B (Reasoning) is Nous Research's reasoning-mode variant of its Llama 3.1 405B fine-tune. The reasoning mode delivers a meaningful uplift in technical benchmarks: livecodebench rises to 0.686 and AIME-25 to 0.697, making it competitive for maths and coding tasks within its tier. GPQA at 0.727 and MMLU-Pro at 0.829 are also respectable.
For businesses, the reasoning variant is the better choice over the non-reasoning sibling for any technical workload. It suits coding assistance, mathematical problem solving, and structured analytical tasks. However, agentic reliability (index 16.8) is low, and instruction following (ifbench 0.327) is only moderate, limiting its use in autonomous pipelines or client-facing content.
At $1/$3 per million tokens it is expensive for its intelligence tier. Teams that specifically need a large open-weight reasoning model and can self-host will get more value; those paying API rates should compare against more capable models at similar or lower prices.
Assessed June 30, 2026
Editorial notes
Hermes 4 405B (Reasoning) from Nous Research shows strong maths and coding benchmarks (livecodebench 0.686, AIME-25 0.697) for a fine-tuned open model, but its intelligence index remains low and pricing is high relative to capability.
Rankings consider pricing, capabilities, benchmarks, and real-world applicability and are refreshed as new models launch. Feedback?
Performance Profile
How Hermes 4 – Llama-3.1 405B (Reasoning) compares
Hermes 4 – Llama-3.1 405B (Reasoning) ranks #247 of 385 AI models we track for overall intelligence, #203 of 293 for agentic tasks. At $1.00 per million input tokens it is cheaper than 25% of comparable models.
Performance Indices
Source: Artificial Analysis
Benchmark Scores
Intelligence
Technical
Content
Benchmark data from Artificial Analysis and Hugging Face
How does Hermes 4 – Llama-3.1 405B (Reasoning) stack up?
Compare side-by-side with other professional models.
Model Information
| Provider | Nous Research |
| Release Date | August 27, 2025 |
| Status | Active |
Pricing
| Token Type | Cost per 1M tokens | Cost per 1K tokens |
|---|---|---|
| Input | $1.00 | $0.001000 |
| Output | $3.00 | $0.003000 |
Leaderboard Categories
Frequently asked questions about Hermes 4 – Llama-3.1 405B (Reasoning)
How much does Hermes 4 – Llama-3.1 405B (Reasoning) cost?
Hermes 4 – Llama-3.1 405B (Reasoning) costs $1.00 per million input tokens and $3.00 per million output tokens.
Who created Hermes 4 – Llama-3.1 405B (Reasoning)?
Hermes 4 – Llama-3.1 405B (Reasoning) is developed by Nous Research and was released on August 27, 2025.
