Nex-N2-Pro
Nex AGI's Nex-N2-Pro is an open-source agentic MoE model with Agentic Thinking, 262K context, tool calling, and coding-heavy workflows for long-horizon use.
Model Overview
Capabilities, design details, and architectural traits
Nex-N2-Pro - Agentic Thinking model built for action
Nex-N2-Pro is an open-source model from Nex AGI that centers on Thinking as a shared pattern across search, coding, and tool calling. It is described as built for action, with adaptive reasoning that can turn on based on task complexity.
| Distinctive trait | What it means |
|---|---|
| Agentic Thinking | The model uses one reasoning structure across search, coding, and agentic tool calling. |
| Adaptive reasoning | It can decide whether to engage Thinking and adjust reasoning depth to the task. |
| Mixed task flow | Its reasoning pattern is described as goal decomposition, state tracking, strategy adjustment, and self-verification. |
| Open-source Nex-N2-Pro | The model is the Pro member of the Nex-N2 family, built on a Qwen3.5-397B-A17B base. |
What it is
Nex-N2-Pro is positioned around long-horizon work, especially coding and tool-driven tasks. The model card and official pages frame it as part of a family that unifies reasoning and action in one workflow.
Core setup
- Architecture: MoE.
- Context: 262K.
- Base: Qwen3.5-397B-A17B.
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 | Nex AGI | Anthropic | Anthropic |
| Release Date | June 2, 2026 | September 22, 2026 | September 28, 2026 |
| Knowledge Cutoff | - | - | Jun 2026 |
| Context & Limits | |||
| Context Window | 262K | 1M Best Context Window | 1M Best Context Window |
| Pricing (per 1M tokens) | |||
| Input Pricing | $0.50 Best Input Pricing | $4 | $2 |
| Output Pricing | $2.50 Best Output Pricing | $20 | $10 |
| Modalities | |||
| Inputs | textimage | textimagefile | textimagefile |
| Outputs | text | text | text |
| Benchmarks (0-100) | |||
| Intelligence Index | 28.2 | 57.6 Best Intelligence Index | 56.0 |
| Coding Index | 59.1 | - | - |
| Agentic Index | 23.7 | - | - |
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.
Top AI Models
Leading alternatives by intelligence score
