Nanbeige LLM Lab根据Apache 2.0许可，在Hugging Face上拥有发布 Nanbeige4.3-33B. Nanbeige4.3-33B是一个33B的密集语言模型，为代理编码和办公室风格的"合作"任务所建。 实验室坐落在招募平台BOSS Zhipin内。 它还为定制后培训出版了"南北"(Nanbeige4.3-33B-Base). 该模型支持524,288个上下文窗口。 在实验室自己的桌子上，它引领了20个基准测试中的17个，与30B附近的3个开放型号相对。
TL; DR 数据
- 大小：33B参数，密度(无MOE),524,288-token(512K)上下文，BF16重量。
多GPU服务器。 该卡的SGLang和vLLM配方使用4向抗声道并列。 BF16重量需要约66GB(我们的算术). 列出1个社区量化。
- 性能：在南碧芝自己与Qwen3.8-27B,Muse-Glimer-30B和Gemma4-31B的对比中领先了20行中的17行。
- 最佳：80.9起终站-奔驰2.1起，Qwen3.8-27B起比73.0起。
- 最差：23.8分，终点站-Bench 4.0分(最低分). 它也跟随了SWE-Bench Pro上的Qwen3.8-27B,60.3对61.7.
南碧盖4.3-33B是什么？
Nanbeige4.3-33B是"南北"4.3系列的旗舰"老师"模式。 实验室的公开策略是首先培养出一位强壮的教师。 随后它计划将代理技能分解为更小的模型，从南比克4.3-3B开始。
Nanbeige4.3-33B如何建设？.
config.json宣布标准LlamaForCausalLM架构。 没有专家的路由，所以所有33B参数都运行在每个符上。
- 52个层，隐藏大小7 168
- 56个查询头和8个关键值头(分组-清点)，头尺寸128
- FFN中间尺寸22 016，词汇166 912个令牌
- 最高职位120 000 000和524 288个
聊天模板曝光3个控件。 启用 thinking 切换推理并默认进行。 保存思考使推理从早期转向。 Nanbeige推荐它用于多回合工具使用并编码为agents. 通过工具可以使工具调用，建议采用XML调用格式。
南碧革4.3-33B在基准测试上表现如何？.
南碧姬将模型与Qwen3.8-27B,Muse-Glimer-30B和Gemma4-31B作比较。 每个模型都以最高的思维能力运行，拥有共享的超参数和杠杆。 本节各分自报。
- 代理编码：它在Terminal-Bench 2.1上得分80.9分，而Qwen3.8-27B则得分73.0分。 DeepSWE 显示最大差距：65.5对42.5. 在Terminal-Bench 4.0上，它得分23.8分，而Gemma4-31B则得分0.0分。
- 代理合作：它把54.2放在自动化座椅上，38.3放在工作本克上，44.8放在工作本克上，34.5放在工作本克。 在YC-Bench上，模拟收入为1 025 437美元，对641 290美元。
- 理由：它报告关于AIME 2026的95.2和关于AA-LCR的84.0。
以二十行相减为三。 Qwen3.8-27B领先SWE-Bench Pro (61.7对60.3)和GDPval (86.4对84.8). Gemma4-31B领先GPQA Diamond出道于南碧杰的桌子，91.3对89.5.
Base模型在所有8个已报任务上都击败了Gemma4-31B-Base. 例如MMLU-Pro(66.9诉59.3)和GPQA(62.0诉41.5)。
它如何与竞争对手自己的样板卡相比较？
* 南比杰自报。 竞争者价值来自每个实验室自己的Hugging Face型号卡。
有些竞争者在南碧芝的表格中
- Meta报告Muse-Glimer-30B在51.2上SWE-Bench Pro和Gemma4-31B在36.9. 南碧洁列出倒数。
- Google在GPQA钻石上列出Gemma 4 31B 分84.3分。 南碧盖的桌子显示 91.3.
- Meta报告Muse-Glimer-30B在Bench 2.1航站楼的51.7站台。 南碧芝测得61.8分。
不同的拉带可以解释一些差距。 SWE-Bench Pro一对看起来是互换的。 在独立跑道出现之前，将这些行视为临时行。
如何部署"南北"4.3-33B?.
模型卡提供变形金刚，SGLang和vLLM配方。 它的服务器实例使用与Qwen3推理和工具调用解析器相平行的4-way收音机。 推荐的 agent 设置为温度 1.0，上 p 0.95 和 65,536 最大新令牌。 理性和聊天使用温度为0.6和262,144的最大新标志。
通过从config.json的算术，BF16重量需要约66GB. KV缓存的新增在BF16中大约为每枚208KiB. 因此，完整的524,288个托肯序列需要大约112克伏快取。 在出版时，没有Hugging Face推论提供者主持模型。
关键外卖
-33B 稠密，Apache 2.0，上下文512K，只用文字。
- 在Bench 2.1号终端站和DeepSWE站站点上自报80.9号终端站和65.5号终端站。
- 在南比吉自己的桌子上领先20行中的17行。
- 几位竞争者人物与官方牌相冲突。
- 建为即将到来的"南北"4.3-3B的老师。
请检查access-date=中的日期值 (帮助) 型号权重，"基地"型号和"X"上"发布"的帖子。 并加入我们的通讯。 等等，你在电报里吗？现在你也可以加入我们的电报。
需要和我们合作推广您的 GitHub Repo OR Hugging Face Page OR Product Release OR Webinar等吗？联系我们
Asif Razzaq是Marktechpost AI Media Inc.的首席执行官。 作为一个有远见的企业家和工程师，阿西夫致力于利用人工智能的潜力为社会造福。 他最近的努力是人工智能媒体平台的发布，该平台以深入报道机器学习和深层学习新闻而出名，这些新闻在技术上既可靠又为广大观众所容易理解。 该平台每月有超过200万次观看，表明其受观众欢迎。
本文由 AI宝库自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
Nanbeige LLM Lab has released Nanbeige4.3-33B on Hugging Face under the Apache 2.0 license. Nanbeige4.3-33B is a 33B dense language model built for agentic coding and office-style ‘cowork’ tasks. The lab sits inside the recruitment platform BOSS Zhipin. It also published Nanbeige4.3-33B-Base for custom post-training. The model supports a 524,288-token context window. In the lab’s own table, it leads 17 of 20 benchmarks against 3 open models near 30B.
TL;DR
- Size: 33B parameters, dense (no MoE), 524,288-token (512K) context, BF16 weights.
- Runs on: Multi-GPU servers. The card’s SGLang and vLLM recipes use 4-way tensor parallelism. BF16 weights need about 66 GB (our arithmetic). 1 community quantization is listed.
- Performance: Leads 17 of 20 rows in Nanbeige’s own comparison against Qwen3.8-27B, Muse-Glimmer-30B and Gemma4-31B.
- Best: 80.9 on Terminal-Bench 2.1, versus 73.0 for Qwen3.8-27B.
- Worst: 23.8 on Terminal-Bench 4.0 (its lowest score). It also trails Qwen3.8-27B on SWE-Bench Pro, 60.3 vs 61.7.
What is Nanbeige4.3-33B?
Nanbeige4.3-33B is the flagship ‘teacher’ model of the Nanbeige4.3 series. The lab’s stated strategy is to build a strong teacher first. It then plans to distill agentic skills into smaller models, starting with Nanbeige4.3-3B.
How is Nanbeige4.3-33B built?
The config.json declares a standard LlamaForCausalLM architecture. There is no expert routing, so all 33B parameters run on every token.
- 52 layers, hidden size 7,168
- 56 query heads and 8 key-value heads (grouped-query attention), head dimension 128
- FFN intermediate size 22,016, vocabulary of 166,912 tokens
- RoPE theta of 120,000,000 and 524,288 max positions
The chat template exposes 3 controls. enable_thinking toggles reasoning and is on by default. preserve_thinking keeps reasoning from earlier turns. Nanbeige recommends it for multi-turn tool use and code agents. Passing tools enables tool calling, and an XML call format is recommended.
How does Nanbeige4.3-33B perform on benchmarks?
Nanbeige compared the model with Qwen3.8-27B, Muse-Glimmer-30B and Gemma4-31B. Every model ran at its highest thinking effort, with shared hyperparameters and harnesses. All scores in this section are self-reported.
- Agentic coding: It scores 80.9 on Terminal-Bench 2.1, against 73.0 for Qwen3.8-27B. DeepSWE shows the widest gap: 65.5 versus 42.5. On Terminal-Bench 4.0, it scores 23.8 while Gemma4-31B scores 0.0.
- Agentic cowork: It posts 54.2 on Automation Bench versus 38.3, and 44.8 on JobBench versus 34.5. On YC-bench, simulated earnings reach $1,025,437 versus $641,290.
- Reasoning: It reports 95.2 on AIME 2026 and 84.0 on AA-LCR.
It loses 3 of 20 rows. Qwen3.8-27B leads SWE-Bench Pro (61.7 vs 60.3) and GDPval (86.4 vs 84.8). Gemma4-31B leads GPQA Diamond in Nanbeige’s table, 91.3 vs 89.5.
The Base model beats Gemma4-31B-Base on all 8 reported tasks. Examples include MMLU-Pro (66.9 vs 59.3) and GPQA (62.0 vs 41.5).
How does it compare with competitors’ own model cards?
*Self-reported by Nanbeige. Competitor values come from each lab’s own Hugging Face model card.
Some competitor figures in Nanbeige’s table differ from what those labs publish:
- Meta reports Muse-Glimmer-30B at 51.2 on SWE-Bench Pro and Gemma4-31B at 36.9. Nanbeige lists the reverse.
- Google lists Gemma 4 31B at 84.3 on GPQA Diamond. Nanbeige’s table shows 91.3.
- Meta reports Muse-Glimmer-30B at 51.7 on Terminal-Bench 2.1. Nanbeige measured 61.8.
Different harnesses explain some gaps. The SWE-Bench Pro pair looks swapped. Treat these rows as provisional until independent runs appear.
How do you deploy Nanbeige4.3-33B?
The model card provides Transformers, SGLang and vLLM recipes. Its server examples use 4-way tensor parallelism with Qwen3 reasoning and tool-call parsers. Recommended agent settings are temperature 1.0, top_p 0.95 and 65,536 max new tokens. Reasoning and chat use temperature 0.6 and 262,144 max new tokens.
By arithmetic from config.json, BF16 weights need about 66 GB. The KV cache adds about 208 KiB per token in BF16. A full 524,288-token sequence therefore needs roughly 112 GB of KV cache. At publication, no Hugging Face Inference Provider hosts the model.
Key Takeaways
- 33B dense, Apache 2.0, 512K context, text-only.
- Self-reported 80.9 on Terminal-Bench 2.1 and 65.5 on DeepSWE.
- Leads 17 of 20 rows in Nanbeige’s own table.
- Several competitor figures conflict with official cards.
- Built as the teacher for the upcoming Nanbeige4.3-3B.
Check out the model weights, the Base model and the launch post on X. Also, feel free to follow us on Twitter and don’t forget to join our 150k+ML SubReddit and Subscribe to our Newsletter. Wait! are you on telegram? now you can join us on telegram as well.
Need to partner with us for promoting your GitHub Repo OR Hugging Face Page OR Product Release OR Webinar etc.? Connect with us
Asif Razzaq is the CEO of Marktechpost AI Media Inc.. As a visionary entrepreneur and engineer, Asif is committed to harnessing the potential of Artificial Intelligence for social good. His most recent endeavor is the launch of an Artificial Intelligence Media Platform, Marktechpost, which stands out for its in-depth coverage of machine learning and deep learning news that is both technically sound and easily understandable by a wide audience. The platform boasts of over 2 million monthly views, illustrating its popularity among audiences.
Auto-fetched and translated by AI Gem Box. All rights belong to the original author.