Model library / nex-agi
Nex-N2.5-mini is a 35.1 billion-parameter mixture-of-experts language model released under Apache-2.0.
Mixture of experts, count not published
Context length in tokens, as published
Weights at the released precision
Licence declared on the repository
Overview
Nex-N2.5-mini is part of the Nex-N2.5 family from Nex-AGI, described as an agentic model for long-horizon tasks that involve computer use, web browsing and visual feedback. It builds on the multimodal foundations of Nex-N2 and adds improvements for autonomous program execution and environment perception.
The model uses a Qwen-3.5 MoE architecture with 35.1 billion parameters and supports a 262,144-token context window. Weights are provided in bfloat16 format and are licensed under Apache-2.0, allowing unrestricted commercial use. The repository does not disclose the number of experts, but the MoE design implies multiple specialised sub-models.
Specification
| Repository | nex-agi/Nex-N2.5-mini |
| Publisher | nex-agi |
| Published | 8 September 2026 |
| Architecture | Qwen3_5MoeForConditionalGeneration |
| Model type | qwen3_5_moe |
| Parameters | 35,107,181,936 |
| Layers | 40 |
| Hidden size | 2,048 |
| Attention heads | 16, 2 key/value heads |
| Context length | 262,144 tokens |
| Vocabulary | 248,320 tokens |
| Weight format | bfloat16 |
| Licence | apache-2.0 |
Memory
Parameter count multiplied by the bytes each format uses. The key-value cache comes on top.
| Precision | Weights | Notes |
|---|---|---|
| bf16 | 65.4 GiB | Full precision as released |
| fp8 | 32.7 GiB | 8-bit, near-lossless on most models |
| int4 | 16.3 GiB | 4-bit, smallest footprint |
The cache costs 80 KiB per token at 16-bit, so the full 262,144-token context of one request needs about 20.0 GiB. Concurrency multiplies that number, not the weights.
Hardware
At the released precision, with 90% of the card given to the server and 2.5 GiB kept for the runtime. The context column is what is left for the cache on that setup.
| GPU | Cards needed | Free for cache | Context it holds |
|---|---|---|---|
| RTX 4090 24GB | 4 | 11.0 GiB | about 144,281 tokens |
| RTX 5090 32GB | 4 | 39.8 GiB | the full 262,144 tokens |
| L40S 48GB | 2 | 16.0 GiB | about 209,817 tokens |
| A100 80GB | 1 | 4.1 GiB | about 53,841 tokens |
| H100 80GB | 1 | 4.1 GiB | about 53,841 tokens |
| RTX PRO 6000 96GB | 1 | 18.5 GiB | about 242,585 tokens |
| H200 141GB | 1 | 59.0 GiB | the full 262,144 tokens |
Serving
A vLLM launch line for the shape above. Check the model card for a runtime the publisher recommends.
vllm serve nex-agi/Nex-N2.5-mini \ --max-model-len 143360 \ --gpu-memory-utilization 0.90
Questions
The weights take about 65 GiB at the released precision. That fits on one A100 80GB, which leaves roughly 4 GiB for the key-value cache.
The repository declares apache-2.0. Read the licence text before commercial use: the name alone does not tell you what is allowed.
The configuration allows 262,144 tokens. Whether the whole window is usable depends on the memory left for the cache, which the hardware table works out per card.
This page is a reference, and Aiqre does not serve this model today. We deploy open-weight models on dedicated EU hardware on request, so if you want an OpenAI-compatible endpoint for it, email [email protected] with the model name and roughly what volume you expect.
Also in the library
Nex-N2.5-Pro is a 396.8 billion-parameter MoE language model designed for long-horizon, agentic tasks with multimodal and compute…
Nex-N2.5-Max is a 1.6-trillion-parameter text-only MoE model designed for long-horizon agentic tasks.
Pantheon-Reasoning-26B-A4B-1.1-V2 is a 26.5 billion-parameter Gemma-4 mixture-of-experts finetune aimed at role-play scenarios, i…
K2-Horizon-32B-Stage1 is a 32-billion-parameter dense decoder-only model released by IFM for benchmarking agentic, coding and rea…
Hosting
Tell us the model and roughly what volume you expect. We reply with a price and an OpenAI-compatible endpoint, with your prompts never stored.