Qwen: Qwen2.5 VL 72B Instruct

Qwen3-30B-A3B-Instruct-2507 is a 30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference. It operates in non-thinking mode and is designed for high-quality instruction following, multilingual understanding, and agentic tool use. Post-trained on instruction data, it demonstrates competitive performance across reasoning (AIME, ZebraLogic), coding (MultiPL-E, LiveCodeBench), and alignment (IFEval, WritingBench) benchmarks. It outperforms its non-instruct variant on subjective and open-ended tasks while retaining strong factual and coding performance.

Qwen3-235B-A22B-Thinking-2507 is a high-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks. It activates 22B of its 235B parameters per forward pass and natively supports up to 262,144 tokens of context. This "thinking-only" variant enhances structured logical reasoning, mathematics, science, and long-form generation, showing strong benchmark performance across AIME, SuperGPQA, LiveCodeBench, and MMLU-Redux. It enforces a special reasoning mode (</think>) and is designed for high-token outputs (up to 81,920 tokens) in challenging domains.

The model is instruction-tuned and excels at step-by-step reasoning, tool use, agentic workflows, and multilingual tasks. This release represents the most capable open-source variant in the Qwen3-235B series, surpassing many closed models in structured reasoning use cases.

Qwen3-Coder-480B-A35B-Instruct is a Mixture-of-Experts (MoE) code generation model developed by the Qwen team. It is optimized for agentic coding tasks such as function calling, tool use, and long-context reasoning over repositories. The model features 480 billion total parameters, with 35 billion active per forward pass (8 out of 160 experts).

Pricing for the Alibaba endpoints varies by context length. Once a request is greater than 128k input tokens, the higher pricing is used.

Qwen3-235B-A22B-Instruct-2507 is a multilingual, instruction-tuned mixture-of-experts language model based on the Qwen3-235B architecture, with 22B active parameters per forward pass. It is optimized for general-purpose text generation, including instruction following, logical reasoning, math, code, and tool usage. The model supports a native 262K context length and does not implement "thinking mode" (<think> blocks).

Compared to its base variant, this version delivers significant gains in knowledge coverage, long-context reasoning, coding benchmarks, and alignment with open-ended tasks. It is particularly strong on multilingual understanding, math reasoning (e.g., AIME, HMMT), and alignment evaluations like Arena-Hard and WritingBench.

Qwen3-0.6B is a lightweight, 0.6 billion parameter language model in the Qwen3 series, offering support for both general-purpose dialogue and structured reasoning through a dual-mode (thinking/non-thinking) architecture. Despite its small size, it supports long contexts up to 32,768 tokens and provides multilingual, tool-use, and instruction-following capabilities.

Qwen3-1.7B is a compact, 1.7 billion parameter dense language model from the Qwen3 series, featuring dual-mode operation for both efficient dialogue (non-thinking) and advanced reasoning (thinking). Despite its small size, it supports 32,768-token contexts and delivers strong multilingual, instruction-following, and agentic capabilities, including tool use and structured output.

Qwen3-4B is a 4 billion parameter dense language model from the Qwen3 series, designed to support both general-purpose and reasoning-intensive tasks. It introduces a dual-mode architecture—thinking and non-thinking—allowing dynamic switching between high-precision logical reasoning and efficient dialogue generation. This makes it well-suited for multi-turn chat, instruction following, and complex agent workflows.

Qwen3, the latest generation in the Qwen large language model series, features both dense and mixture-of-experts (MoE) architectures to excel in reasoning, multilingual support, and advanced agent tasks. Its unique ability to switch seamlessly between a thinking mode for complex reasoning and a non-thinking mode for efficient dialogue ensures versatile, high-quality performance.

Significantly outperforming prior models like QwQ and Qwen2.5, Qwen3 delivers superior mathematics, coding, commonsense reasoning, creative writing, and interactive dialogue capabilities. The Qwen3-30B-A3B variant includes 30.5 billion parameters (3.3 billion activated), 48 layers, 128 experts (8 activated per task), and supports up to 131K token contexts with YaRN, setting a new standard among open-source models.

Qwen3-8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue. It supports seamless switching between "thinking" mode for math, coding, and logical inference, and "non-thinking" mode for general conversation. The model is fine-tuned for instruction-following, agent integration, creative writing, and multilingual use across 100+ languages and dialects. It natively supports a 32K token context window and can extend to 131K tokens with YaRN scaling.

Qwen3-14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue. It supports seamless switching between a "thinking" mode for tasks like math, programming, and logical inference, and a "non-thinking" mode for general-purpose conversation. The model is fine-tuned for instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.

Qwen3-32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue. It supports seamless switching between a "thinking" mode for tasks like math, coding, and logical inference, and a "non-thinking" mode for faster, general-purpose conversation. The model demonstrates strong performance in instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.

Qwen3-235B-A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass. It supports seamless switching between a "thinking" mode for complex reasoning, math, and code tasks, and a "non-thinking" mode for general conversational efficiency. The model demonstrates strong reasoning ability, multilingual support (100+ languages and dialects), advanced instruction-following, and agent tool-calling capabilities. It natively handles a 32K token context window and extends up to 131K tokens using YaRN-based scaling.

Qwen2.5-Coder-7B-Instruct is a 7B parameter instruction-tuned language model optimized for code-related tasks such as code generation, reasoning, and bug fixing. Based on the Qwen2.5 architecture, it incorporates enhancements like RoPE, SwiGLU, RMSNorm, and GQA attention with support for up to 128K tokens using YaRN-based extrapolation. It is trained on a large corpus of source code, synthetic data, and text-code grounding, providing robust performance across programming languages and agentic coding workflows.

This model is part of the Qwen2.5-Coder family and offers strong compatibility with tools like vLLM for efficient deployment. Released under the Apache 2.0 license.

Qwen2.5 VL 3B is a multimodal LLM from the Qwen Team with the following key enhancements:

For more details, see this blog post and GitHub repo.

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen2.5-VL-32B is a multimodal vision-language model fine-tuned through reinforcement learning for enhanced mathematical reasoning, structured outputs, and visual problem-solving capabilities. It excels at visual analysis tasks, including object recognition, textual interpretation within images, and precise event localization in extended videos. Qwen2.5-VL-32B demonstrates state-of-the-art performance across multimodal benchmarks such as MMMU, MathVista, and VideoMME, while maintaining strong reasoning and clarity in text-based tasks like MMLU, mathematical problem-solving, and code generation.

QwQ is the reasoning model of the Qwen series. Compared with conventional instruction-tuned models, QwQ, which is capable of thinking and reasoning, can achieve significantly enhanced performance in downstream tasks, especially hard problems. QwQ-32B is the medium-sized reasoning model, which is capable of achieving competitive performance against state-of-the-art reasoning models, e.g., DeepSeek-R1, o1-mini.

Qwen2.5 32B Instruct is the instruction-tuned variant of the latest Qwen large language model series. It provides enhanced instruction-following capabilities, improved proficiency in coding and mathematical reasoning, and robust handling of structured data and outputs such as JSON. It supports long-context processing up to 128K tokens and multilingual tasks across 29+ languages. The model has 32.5 billion parameters, 64 layers, and utilizes an advanced transformer architecture with RoPE, SwiGLU, RMSNorm, and Attention QKV bias.

For more details, please refer to the Qwen2.5 Blog .

Qwen's Enhanced Large Visual Language Model. Significantly upgraded for detailed recognition capabilities and text recognition abilities, supporting ultra-high pixel resolutions up to millions of pixels and extreme aspect ratios for image input. It delivers significant performance across a broad range of visual tasks.

Qwen VL Max is a visual understanding model with 7500 tokens context length. It excels in delivering optimal performance for a broader spectrum of complex tasks.

Qwen-Turbo, based on Qwen2.5, is a 1M context model that provides fast speed and low cost, suitable for simple tasks.

Qwen-Plus, based on the Qwen2.5 foundation model, is a 131K context model with a balanced performance, speed, and cost combination.

Qwen-Max, based on Qwen2.5, provides the best inference performance among Qwen models, especially for complex multi-step tasks. It's a large-scale MoE model that has been pretrained on over 20 trillion tokens and further post-trained with curated Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF) methodologies. The parameter count is unknown.

QwQ-32B-Preview is an experimental research model focused on AI reasoning capabilities developed by the Qwen Team. As a preview release, it demonstrates promising analytical abilities while having several important limitations:

Qwen2.5-Coder is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen). Qwen2.5-Coder brings the following improvements upon CodeQwen1.5:

To read more about its evaluation results, check out Qwen 2.5 Coder's blog.

Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen2.5 72B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2:

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen2.5 VL 7B is a multimodal LLM from the Qwen Team with the following key enhancements:

For more details, see this blog post and GitHub repo.

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen2 7B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.

It features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.

For more details, see this blog post and GitHub repo.

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen2 72B is a transformer-based model that excels in language understanding, multilingual capabilities, coding, mathematics, and reasoning.

It features SwiGLU activation, attention QKV bias, and group query attention. It is pretrained on extensive data with supervised finetuning and direct preference optimization.

For more details, see this blog post and GitHub repo.

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen1.5 110B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:

For more details, see this blog post and GitHub repo.

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen1.5 72B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:

For more details, see this blog post and GitHub repo.

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen1.5 32B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:

For more details, see this blog post and GitHub repo.

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen1.5 14B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:

For more details, see this blog post and GitHub repo.

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen1.5 7B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:

For more details, see this blog post and GitHub repo.

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.

Qwen1.5 4B is the beta version of Qwen2, a transformer-based decoder-only language model pretrained on a large amount of data. In comparison with the previous released Qwen, the improvements include:

For more details, see this blog post and GitHub repo.

Usage of this model is subject to Tongyi Qianwen LICENSE AGREEMENT.