Third-Party Large Models

Specifications of Third-Party Large Models

In addition to Pangu models, ModelArts Studio integrates popular open-source third-party NLP models.

For example, DeepSeek V3 was released on December 26, 2024. It is a Mixture-of-Experts (MoE) language model with 671B parameters. DeepSeek V3 outperforms GPT-4.5 on mathematical and coding evaluation benchmarks. DeepSeek R1 has a similar structure to DeepSeek V3. It was officially open-sourced on January 20, 2025. As an outstanding representative of models with strong reasoning capabilities, DeepSeek R1 has attracted great attention. DeepSeek R1 has achieved the same or even better performance than top closed-source models such as GPT-4o and GPT-4o1 in core tasks such as mathematical reasoning and code generation, and is recognized as a leading LLM in the industry.

ModelArts Studio provides you with third-party NLP models of different specifications to meet different scenarios and requirements. The following table lists the supported models. You can choose the most suitable model based on your requirements for development and application.

Supported Region	Model Name	Maximum Context Length	Maximum Output Length	Description
CN-Hong Kong	DeepSeek-R1-32K-0.0.2	32K	8K	It was released in June 2025. It supports inference for a context length of 32K tokens. 16 inference units are required to deploy the model. Inference with a context length of 32K tokens supports up to 256 concurrent calls. The base model of this version is the open-source model DeepSeek R1-0528.
	DeepSeek-V3-32K-0.0.2	32K	8K	It was released in June 2025. It supports inference for a context length of 32K tokens. 16 inference units are required to deploy the model. Inference with a context length of 32K tokens supports up to 256 concurrent calls. The base model of this version is the open-source model DeepSeek V3-0324.
	DeepSeek-R1-Distil-Qwen-32B-0.0.1	32K	8K	DeepSeek-R1-Distill-Qwen-32B is a model fine-tuned based on the open-source model Qwen2.5-32B using data generated by DeepSeek-R1.
	DeepSeek-R1-Distill-LLama-70B-0.0.1	32K	8K	DeepSeek-R1-Distill-Llama-70B is a model fine-tuned based on the open-source model Llama-3.1-70B using data generated by DeepSeek-R1.
	DeepSeek-R1-Distill-LLama-8B-0.0.1	32K	8K	DeepSeek-R1-Distill-Llama-8B is a model fine-tuned based on the open-source model Llama-3.1-8B using data generated by DeepSeek-R2.
	Qwen3-235B-A22B-0.0.1	32K	8K	Qwen3-235B-A22B uniquely supports seamless switching between thinking mode and non-thinking mode, allowing users to switch between the two in a dialogue. The model's inference capability significantly outperforms that of QwQ, and its general capability far exceeds that of Qwen2.5-72B-Instruct, achieving SOTA performance among models of the same scale in the industry.
	Qwen3-32B-0.0.1	32K	8K	Qwen3-32B uniquely supports seamless switching between thinking mode and non-thinking mode, allowing users to switch between the two in a dialogue. The model's inference capability significantly outperforms that of QwQ, and its general capability far exceeds that of Qwen2.5-32B-Instruct, achieving SOTA performance among models of the same scale in the industry.
	Qwen3-30B-A3B-0.0.1	32K	8K	Qwen3-30B-A3B uniquely supports seamless switching between thinking mode and non-thinking mode, allowing users to switch between the two in a dialogue. The model's inference capability significantly outperforms that of QwQ, and its general capability far exceeds that of Qwen2.5-32B-Instruct, achieving SOTA performance among models of the same scale in the industry.
	Qwen3-14B-0.0.1	32K	8K	Qwen3-14B uniquely supports seamless switching between thinking mode and and non-thinking mode, allowing users to switch between the two in a dialogue. The model's inference capability reaches the SOTA level among models of the same scale in the industry, and its general capability significantly surpasses that of Qwen2.5-14B.
	Qwen3-8B-0.0.1	32K	8K	Qwen3-8B uniquely supports seamless switching between thinking mode and non-thinking mode, allowing users to switch between the two in a dialogue. The model's inference capability reaches the SOTA level among models of the same scale in the industry, and its general capability significantly surpasses that of Qwen2.5-7B.
	Qwen2.5-72B-0.0.1	32K	8K	Compared to Qwen2, Qwen2.5 has acquired significantly more knowledge and has greatly improved capabilities in coding and mathematics. Additionally, the new models achieve significant improvements in instruction following, generating long text, understanding structured data (e.g, tables), and generating structured outputs especially JSON.
	Qwen-QWQ-32B-0.0.1	32K	8K	This model is the QwQ reasoning model trained based on Qwen2.5-32B. Reinforcement learning greatly improves the model's inference capability. The core metrics of the model, including mathematical code (AIME 24/25, LiveCodeBench) and some general metrics (IFEval, LiveBench, etc.), reach the level of the full version of DeepSeek-R1, with all metrics significantly surpassing those of DeepSeek-R1-Distill-Qwen-32B, which is also based on Qwen2.5-32B.

Supported Platform Operations

**Table 1** Platform operations supported by third-party large models
Model Name	Model Evaluation	Real-Time Inference	Model Commissioning in Experience Center
DeepSeek-V3-32K-0.0.2	√	√	√
DeepSeek-R1-32K-0.0.2	√	√	√
DeepSeek-R1-Distil-Qwen-32B-0.0.1	√	√	√
DeepSeek-R1-Distill-LLama-70B-0.0.1	√	√	√
DeepSeek-R1-Distill-LLama-8B-0.0.1	√	√	√
Qwen3-235B-A22B-0.0.1	√	√	√
Qwen3-32B-0.0.1	√	√	√
Qwen3-30B-A3B-0.0.1	√	√	√
Qwen3-14B-0.0.1	√	√	√
Qwen3-8B-0.0.1	√	√	√
Qwen2.5-72B-0.0.1	√	√	√
Qwen-QWQ-32B-0.0.1	√	√	√

Dependency of Third-Party Large Models on Resource Pools

**Table 2** Dependency of third-party large models on resource pools
Model Name	Cloud-based Deployment
Model Name	Arm+Snt9B3
DeepSeek-V3-32K-0.0.2	Supported
DeepSeek-R1-32K-0.0.2	Supported
DeepSeek-R1-Distil-Qwen-32B-0.0.1	Supported
DeepSeek-R1-Distill-LLama-70B-0.0.1	Supported
DeepSeek-R1-Distill-LLama-8B-0.0.1	Supported
Qwen3-235B-A22B-0.0.1	Supported
Qwen3-32B-0.0.1	Supported
Qwen3-30B-A3B-0.0.1	Supported
Qwen3-14B-0.0.1	Supported
Qwen3-8B-0.0.1	Supported
Qwen2.5-72B-0.0.1	Supported
Qwen-QWQ-32B-0.0.1	Supported