
🤗 Hugging Face | 🤖 ModelScope
引言
隆重推出 Ring-2.5-1T:全球首个基于混合线性注意力架构的开源万亿参数思考型模型。
在迈向通用 AI 智能体的重要一步中,我们正在将混合线性注意力扩展至预训练和 RL 阶段。我们高效的 1:7 MLA + Lightning Linear Attention 提升了推理速度与探索能力,而扩展后的 RL 训练则增强了深度思考与长时程任务执行能力。
性能
Ring-2.5-1T 模型在 IMO 2025 和 CMO 2025 上均达到金牌🏅水平的表现。关于我们模型的详细解题过程,请参见示例。
模型下载
| 模型 | 上下文长度 | 下载 |
|---|---|---|
| Ring-2.5-1T | 128K -> 256K (YaRN) | 🤗 HuggingFace 🤖 ModelScope |
注意:如果你对之前的版本感兴趣,请访问 Huggingface 或 ModelScope 中过往的模型合集。
部署
SGLang
环境准备
我们后续会将模型提交至 SGLang 官方发布,现在我们可以按照以下步骤准备环境:
git clone -b ling_2_5 git@github.com:antgroup/sglang.git
cd sglang
# Install the python packages
pip install --upgrade pip
pip install -e "python"
运行推理
SGLang 目前同时支持 BF16 和 FP8 模型。这取决于 ${MODEL_PATH} 中模型的 dtype。
以下是在多 GPU 节点上运行 Ring-1T 的示例,其中主节点 IP 为 ${MASTER_IP},服务器端口为 ${PORT}:
- 启动服务器:
# Node 0:
python -m sglang.launch_server --model-path $MODEL_PATH --tp-size 8 --pp-size 4 --dp-size 1 --trust-remote-code --dist-init-addr $MASTER_IP:2345 --port $PORT --nnodes 4 --node-rank 0
# Node 1:
python -m sglang.launch_server --model-path $MODEL_PATH --tp-size 8 --pp-size 4 --dp-size 1 --trust-remote-code --dist-init-addr $MASTER_IP:2345 --port $PORT --nnodes 4 --node-rank 1
# Node 2:
python -m sglang.launch_server --model-path $MODEL_PATH --tp-size 8 --pp-size 4 --dp-size 1 --trust-remote-code --dist-init-addr $MASTER_IP:2345 --port $PORT --nnodes 4 --node-rank 2
# Node 3:
python -m sglang.launch_server --model-path $MODEL_PATH --tp-size 8 --pp-size 4 --dp-size 1 --trust-remote-code --dist-init-addr $MASTER_IP:2345 --port $PORT --nnodes 4 --node-rank 3
# This is only an example. Please adjust arguments according to your actual environment.
- 客户端:
curl -s http://${MASTER_IP}:${PORT}/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{"model": "auto", "messages": [{"role": "user", "content": "What is the capital of France?"}]}'
更多用法可参见此处
许可证
本代码仓库依据MIT 许可证授权。

🤗 Hugging Face | 🤖 ModelScope
Introduction
Introducing Ring-2.5-1T: the world's first open-source trillion-parameter thinking model based on hybrid linear attention architecture.
In a major step toward general-purpose AI agents, we're scaling hybrid linear attention across pre-training and RL. Our efficient 1:7 MLA + Lightning Linear Attention boosts reasoning speed and exploration, while expanded RL training enhances deep thinking and long-horizon task execution.
Performance
Ring-2.5-1T model achieves gold-medal🏅 level performance at both IMO 2025 and CMO 2025. For detailed solutions of our model, please see examples.
Model Downloads
| Model | Context Length | Download |
|---|---|---|
| Ring-2.5-1T | 128K -> 256K (YaRN) | 🤗 HuggingFace 🤖 ModelScope |
Note: If you are interested in previous version, please visit the past model collections in Huggingface or ModelScope.
Deployment
SGLang
Environment Preparation
We will later submit our model to SGLang official release, now we can prepare the environment following steps:
git clone -b ling_2_5 git@github.com:antgroup/sglang.git
cd sglang
# Install the python packages
pip install --upgrade pip
pip install -e "python"
Run Inference
Both BF16 and FP8 models are supported by SGLang now. It depends on the dtype of the model in ${MODEL_PATH}.
Here is the example to run Ring-1T with multiple GPU nodes, where the master node IP is ${MASTER_IP} and server port is ${PORT}:
- Start server:
# Node 0:
python -m sglang.launch_server --model-path $MODEL_PATH --tp-size 8 --pp-size 4 --dp-size 1 --trust-remote-code --dist-init-addr $MASTER_IP:2345 --port $PORT --nnodes 4 --node-rank 0
# Node 1:
python -m sglang.launch_server --model-path $MODEL_PATH --tp-size 8 --pp-size 4 --dp-size 1 --trust-remote-code --dist-init-addr $MASTER_IP:2345 --port $PORT --nnodes 4 --node-rank 1
# Node 2:
python -m sglang.launch_server --model-path $MODEL_PATH --tp-size 8 --pp-size 4 --dp-size 1 --trust-remote-code --dist-init-addr $MASTER_IP:2345 --port $PORT --nnodes 4 --node-rank 2
# Node 3:
python -m sglang.launch_server --model-path $MODEL_PATH --tp-size 8 --pp-size 4 --dp-size 1 --trust-remote-code --dist-init-addr $MASTER_IP:2345 --port $PORT --nnodes 4 --node-rank 3
# This is only an example. Please adjust arguments according to your actual environment.
- Client:
curl -s http://${MASTER_IP}:${PORT}/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{"model": "auto", "messages": [{"role": "user", "content": "What is the capital of France?"}]}'
More usage can be found here
License
This code repository is licensed under the MIT License.