Hermes 4 70B
Hermes 4 70B from Nous Research is a hybrid reasoning model with `<think>` traces, 131K context, JSON schema adherence, tool use, and reduced refusals.
Model Overview
Capabilities, design details, and architectural traits
Hermes 4 70B - Hybrid Reasoning Model
Hermes 4 70B is a frontier, hybrid-mode reasoning model from Nous Research based on Llama-3.1-70B. It is positioned around deliberate reasoning, structured outputs, and developer-controlled interaction patterns.
Defining Characteristics
| Aspect | Official Characteristic |
|---|---|
| Reasoning mode | Can generate explicit` deliberation traces before answering. |
| Interaction control | Reasoning behavior can be enabled with thinking=True or the reasoning.enabled boolean. |
| Context strategy | Supports a 131,072-token context window. |
| Structured output | Trained to produce valid JSON for schemas and repair malformed objects. |
| Tool use | Supports function/tool calls within a single assistant turn. |
| Alignment profile | Described as easier to steer, with reduced refusal rates and broadly neutral alignment. |
Benchmark Performance
Independent evaluations · Artificial Analysis
Accuracy & Capability Details
Compare Models Side-by-Side
Evaluate specifications, pricing, and independent benchmark indices
| Model Details | |||
|---|---|---|---|
| General Info | |||
| Provider | Nous Research | Anthropic | Anthropic |
| Release Date | August 27, 2025 | September 22, 2026 | September 28, 2026 |
| Knowledge Cutoff | Aug 2024 | - | Jun 2026 |
| Context & Limits | |||
| Context Window | 131K | 1M Best Context Window | 1M Best Context Window |
| Pricing (per 1M tokens) | |||
| Input Pricing | $0.13 Best Input Pricing | $4 | $2 |
| Output Pricing | $0.40 Best Output Pricing | $20 | $10 |
| Modalities | |||
| Inputs | text | textimagefile | textimagefile |
| Outputs | text | text | text |
| Benchmarks (0-100) | |||
| Intelligence Index | 6.7 | 57.6 Best Intelligence Index | 56.0 |
| Coding Index | - | - | - |
| Agentic Index | - | - | - |
Humanity's Last Exam
Extremely difficult logical reasoning and knowledge.
Long Context Reasoning
Logical reasoning over long context windows.
SciCode Benchmark
Scientific coding and mathematical modeling.
Independent evaluation data provided by Artificial Analysis. To view the latest benchmarks and full details, visit their official site.
Explore more from Nous Research
Other models by Nous Research
Top AI Models
Leading alternatives by intelligence score
