Vendor · Language models
Nex AGI has 5 models in the directory: 5 language models.
In the directory: 5 language models. 5 of them with open weights that can run locally.
Every model there ever was. Models with a profile lead to every benchmark, price and source.
| Nex-N2-mini | Jun 4, 2026 | Active | n/a | n/a | n/a | 262K | 34 |
| Nex-N2-Pro | Jun 3, 2026 | Active | n/a | n/a | 117 | 262K | 27 |
| Nex-N2.5-Max | n/a | Open weights | n/a | n/a | n/a | n/a | below 1 |
| Nex-N2.5-mini | n/a | Open weights | n/a | n/a | n/a | n/a | below 1 |
| Nex-N2.5-Pro | n/a | Open weights | n/a | n/a | n/a | n/a | below 1 |
Language model · Jun 4, 2026
Nex-N2-mini combines 35 billion total parameters with about 3 billion active and Nex AGI's agentic thinking approach. Built on Qwen 3.5 35B A3B, its Adaptive Thinking stays brief on simple actions and works more deeply on critical decisions. Coherent Thinking is intended to keep that shift consistent across general and agentic tasks.
Language model · Jun 3, 2026
Nex-N2-Pro combines 397 billion total parameters with about 17 billion active on top of Qwen 3.5 397B A17B. Adaptive Thinking controls reasoning depth, while Coherent Thinking is meant to keep the derivation consistent across longer workflows. Tool calling and explicit reasoning traces target demanding work chains.
Language model · n/a
Nex-N2.5-Max marks Nex-AGI’s first complete post-training effort at trillion-parameter scale, according to the company. Its foundation is a text-only MoE model with 1.6 trillion parameters. Unlike Mini and Pro, Max does not accept visual input.
Language model · n/a
Nex-N2.5-mini uses visual feedback to work through longer computer and browser tasks. Nex-AGI trains this compact multimodal variant on a broader set of environments and tasks than the previous Nex-N2 series.
Language model · n/a
Nex-N2.5-Pro extends the visual computer and browser capabilities of the Nex-N2 series. The model can inspect intermediate results through visual feedback. In Nex-AGI’s own OSWorld-2 evaluation, Pro scores 56.4%, compared with 30.5% for Mini.
| Vendor | Models | Model types | Best language model | Top 10 | Open weights | New in 2026 |
|---|---|---|---|---|---|---|
| Anthropic | 33 | 33 language models | Claude Opus 5.5 (Rank 1 to 3) | 4 | 0 | 13 |
| OpenAI | 63 | 46 language models, 4 image models, 2 video models, 9 audio models, 2 specialist models | GPT-6 Astra (Rank 4 to 8) | 2 | 4 | 27 |
| 66 | 43 language models, 4 image models, 6 video models, 11 audio models, 2 specialist models | Gemini 4 Carbon (Rank 1 to 5) | 2 | 19 | 26 | |
| Meta | 32 | 29 language models, 1 image model, 1 video model, 1 specialist model | Muse Wonder Full (Rank 6 to 9) | 2 | 26 | 7 |
| DeepSeek | 23 | 23 language models | DeepSeek V4.1 Flash 1017 (Rank 13 to 20) | 0 | 23 | 9 |
| Moonshot AI | 12 | 12 language models | Kimi K3.1 Pro (Rank 10 to 13) | 0 | 10 | 7 |
| xAI | 16 | 12 language models, 2 image models, 2 video models | Grok 4.7 (Rank 18 to 24) | 0 | 1 | 12 |
| Nex AGI | 5 | 5 language models | not ranked yet | 0 | 5 | 2 |
By release date. Grey: retired models.
2 releases since January 2026, by release date.
The directory lists 5 Nex AGI models: 5 language models. It counts every model Nex AGI ever released, including retired ones. 5 are currently available, the newest being Nex-N2-mini, Nex-N2-Pro and Nex-N2.5-Max.
Nex-N2-mini, released on Jun 4, 2026. The timeline below shows every release by date.
Yes, 5 of the 5 models have open weights. Each model profile lists how much memory it needs.
Nex-N2-Pro with 117 tokens per second according to Artificial Analysis. A typical model manages 98.
Up to 262,144 tokens (≈ 390 A4 pages), which 2 of the current models reach, including Nex-N2-mini. Every model's value is in the model list.
Data as of Nov 1, 2026