Qwen
ChinaReported upcoming Qwen speech recognition model.
Reported upcoming Qwen realtime audio model.
Qwen's officially documented multilingual speech synthesis model with streaming, expressive speech controls and voice cloning.
Qwen Image 2.1 is an open-weight model for text-to-image generation and image editing. Its visual generation component has 7B parameters, with native transparent RGBA generation and editing, local edit controls, and support for up to 10 reference images.
Qwen's multimodal understanding model accepting text, image, audio and video input with text output, reasoning, tool calling and a million-token context window.
Qwen3.8-Max-0902 is a pinned Qwen Cloud snapshot of Qwen3.8-Max, further post-trained for coding and collaborative agent work. It improves engineering-scale coding, long-horizon autonomous development, multi-tool orchestration, enterprise tasks, scientific research, chart reasoning, document parsing and multimodal perception while retaining the 1M-token context window, thinking mode and built-in tool ecosystem. It is also available under the alias qwen3.8-max-2026-09-02.
Qwen 3.8 Flash is Qwen's fast multimodal model for reasoning, coding assistance, agentic workflows, visual understanding, and long-context applications. Model Studio advertises a million-token context window and OpenAI- and Anthropic-compatible APIs.
Wan 3.0 Video Prime is the speed-optimized Alibaba Cloud Model Studio Wan 3.0 video generation model. It keeps the standard model's all-in-one text-to-video, first/last-frame, multimodal reference, video editing, video extension, and native audio capabilities while improving end-to-end generation speed.
Qwen 3.8 27B is a dense 27.8-billion-parameter open-weight vision-language model from the Qwen Team at Alibaba. It supports text, image and video inputs, a 262,144-token native context window extendable to 1 million tokens, tool use and controllable reasoning.
Qwen3.8-2.4T-A95B is Qwen's open-weight, text-only mixture-of-experts model for coding, professional work, research and long-horizon agentic tasks. It has 2.4 trillion total parameters with 95 billion active per token, configurable reasoning effort, a native 262,144-token context window and support for extending context to 1,010,000 tokens. Qwen 3.8 Max is a separate managed API model based on this checkpoint with additional multimodal and hosted features.
Wan 3.0 Video is Alibaba Cloud Model Studio's all-in-one video generation model for text-to-video, first/last-frame image-to-video, multimodal reference generation, video editing, and video extension. It generates 30 fps MP4 video with native dialogue, background music, and sound effects.
Qwen 3.8 Max is Qwen Cloud's managed multimodal flagship for complex reasoning, coding and professional work. It is based on Qwen3.8-2.4T-A95B but is a distinct API model with native text, image and video understanding, non-thinking support, a default 1M-token context window, function calling, structured outputs, caching, batches and built-in tools.
Qwen3.7 Flash is a multimodal Qwen model optimized for efficient generation.
Qwen3.8-Max-Preview is a 2.4T-parameter preview model from Qwen. It has debuted through Alibaba Token Plan, Qoder, and QoderWork while Qwen continues testing it in China; its architecture, context limits, public API model ID, pricing, and benchmark results have not been published.
Qwen 3.7 Plus is the stable Qwen 3.7 Plus release for multimodal reasoning, coding, tool use, and productivity workflows. It accepts text, image, and video input, produces text output, and supports cache, function calling, prefix completion, structured outputs, and web search.
Qwen 3.7 Plus (2026-05-26) is a dated Qwen 3.7 Plus snapshot for multimodal reasoning, coding, tool use, and productivity workflows. It accepts text, image, and video input, produces text output, and reflects the May 26, 2026 snapshot exposed by Alibaba Cloud.
Qwen 3.7 Max (2026-05-17) is a dated Qwen 3.7 Max snapshot for reasoning and chat workloads. It accepts text-only input, enables thinking mode by default, and does not expose the cache features available on the stable 2026-05-20 release.
Qwen 3.7 Max Preview is a preview Qwen text-generation model for reasoning and chat workloads. It accepts text-only input, enables thinking mode by default, and does not expose the cache features available on the stable 2026-05-20 release.
Qwen 3.7 Max is the stable Qwen 3.7 Max release backed by the 2026-05-20 version. It accepts text-only input, enables thinking mode by default, and supports explicit, implicit, and session cache.
Qwen 3.6 27B is a dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026. It accepts text and image inputs and produces text outputs.
Qwen 3.5 Plus (2026-04-20) is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window.
A fast, cost-efficient Qwen 3.6 model with a one-million-token context window.
Qwen 3.6 35B A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per token. It supports turning still images into animated clips as well as prompt-driven video workflows.
Wan 2.7 T2V is a video generation model from Qwen for prompt-driven clip and scene creation. It is aimed at text-to-video, multimodal clip generation, and motion design workflows.
Qwen 3.6 Plus builds on a hybrid architecture that combines efficient linear attention with sparse mixture-of-experts routing, enabling strong scalability and high-performance inference.
Qwen 3.5 0.8B is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video.
Qwen 3.5 2B is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video. It accepts text and image inputs and produces text outputs.
Qwen 3.5 4B is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video. It accepts text and image inputs and produces text outputs.
Qwen 3.5 9B is a multimodal foundation model from the Qwen3.5 family, designed to deliver strong reasoning, coding, and visual understanding in an efficient 9B-parameter architecture.
The Qwen3.5 native vision-language Flash models are built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. Compared to the...
Qwen 3.5 122B A10B is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency.
The Qwen3.5 27B native vision-language Dense model incorporates a linear attention mechanism, delivering fast response times while balancing inference speed and performance.
Qwen 3.5 35B A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inference efficiency.
Qwen 3.5 Flash is a fast, efficient language model from Alibaba's Qwen 3.6 series. It supports text, image, and video input with a 1M token context window.
The Qwen3.5 native vision-language series Plus models are built on a hybrid architecture that integrates linear attention mechanisms with sparse mixture-of-experts models, achieving higher inference efficiency. In a variety of...
Qwen 3.5 397B A17B is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency.
Qwen 3.5 Plus is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window.
Qwen 3.5 Plus (2026-02-15) is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window.
Qwen 3 Max Thinking is the flagship reasoning model in the Qwen3 series, designed for high-stakes cognitive tasks that require deep, multi-step reasoning.
Qwen3Guard-Gen 0.6B is a generative safety classifier that labels content Safe, Controversial, or Unsafe and reports safety categories and a refusal decision.
Qwen3Guard-Gen 8B is a generative safety classifier that labels content Safe, Controversial, or Unsafe and reports safety categories and a refusal decision.
Qwen3 LiveTranslate Flash Realtime is a low-latency multilingual interpretation model for real-time audio and video translation workflows. It accepts audio, text, image, and video inputs and can return translated audio and text outputs.
Qwen3-Coder-30B-A3B-Instruct is a 30.5B parameter Mixture-of-Experts (MoE) model with 128 experts (8 active per forward pass), designed for advanced code generation, repository-scale understanding, and agentic tool use. Built on the...
Qwen3-Coder-480B-A35B-Instruct is the Qwen3's most agentic code model, featuring Significant Performance on Agentic Coding, Agentic Browser-Use and other foundational coding tasks, achieving results comparable to Claude Sonnet.
Qwen3-235B-A22B-Instruct-2507 is the updated version of the Qwen3-235B-A22B non-thinking mode, featuring Significant improvements in general capabilities, including instruction following, logical reasoning, text comprehension, mathematics, science, coding and tool usage.
Qwen 3 Embedding 8B is a multilingual text embedding model for semantic search, retrieval, clustering, classification, and RAG workflows.
Qwen 2.5 VL 32B Instruct is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Qwen 2.5 VL 72B Instruct
Qwen3, a 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.
Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Qwen 2.5 VL 72B Instruct is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
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.
Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2: - Significantly more knowledge and has greatly improved capabilities in coding and...
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.
Qwen 3.8 Flash Next is an open-weight multimodal model served by RunInfra with text and image input, tool calling, JSON mode, streaming, and a 1,048,576-token context window.
Qwen3.8-27B is Alibaba's open-weight 27B dense model, released under Apache-2.0 and self-hosted on OrcaRouter's own infrastructure. This deployment is text-in, text-out, with a 64K-token context window (65,536 positions) built on the Qwen3.5 architecture - 64 layers, 5120 hidden size. Despite its 27B size it posts unusually strong agentic and coding results on Qwen's own model card, including 90.3 on LiveCodeBench v6, 89.2 on GPQA Diamond, and 79.0 on QwenSWEBench - competitive with far larger closed models on several axes. It supports native tool calling, structured outputs, reasoning mode, and the full sampling surface. Because the weights are open and we run them ourselves, there is no per-token vendor cost to pass through. That makes Qwen3.8-27B a strong default for high-volume text workloads: coding agents, long-document analysis, structured extraction, and agentic automation.
Qwen3.5 Flash snapshot 2026-02-23 — frozen version of qwen3.5-flash, same capabilities.
Qwen3.6 Flash snapshot 2026-04-16 — frozen version of qwen3.6-flash, same capabilities.
Qwen3.6 Plus snapshot 2026-04-02 — frozen version of qwen3.6-plus, same capabilities.
Qwen3.7-Max (2026-05-20 snapshot) — Dated checkpoint of Alibaba's flagship proprietary agent-era model, pinned for reproducible production workloads. Native 1M token context window, with an extended thinking mode (and preserve_thinking across turns) tuned for agentic tasks. Frontier-level results on coding (SWE-Verified, SWE-Pro, Terminal-Bench), reasoning (GPQA Diamond, HMMT, IMO), tool use (BFCL, MCP-Mark, MCP-Atlas), and multilingual benchmarks (WMT24++ across 55 languages). Engineered for long-horizon autonomous execution and consistent behavior across agent scaffolds including Claude Code, OpenClaw, and Qwen Code. Use this pinned version when you need stable behavior across releases; use qwen/qwen3.7-max for the rolling alias.
Qvq Max is a video generation model from Qwen for prompt-driven clip and scene creation. It supports turning still images into animated clips as well as prompt-driven video workflows.
Qwen 3 Reranker 8B rescoring model for ranking text documents against a text query.
Qwen 3 TTS (1.7B) is a DigitalOcean-hosted text-to-speech model exposed through DigitalOcean Inference.
Qwen 3.6 Max Preview is a proprietary frontier model from Alibaba Cloud built on a sparse mixture-of-experts architecture with approximately 1 trillion total parameters.
Qwen 3.7 Plus Preview is a proprietary preview model entry from Qwen for fast, general-purpose multimodal assistance.
Qwen Flash is a language model from Qwen for chat, writing, analysis, and general-purpose automation. It is tuned for lower latency and cost-sensitive production use.
Qwen Flash Character is a language model from Qwen for chat, writing, analysis, and general-purpose automation. It is tuned for lower latency and cost-sensitive production use.
Qwen Image Edit Max is a proprietary frontier model from Alibaba Cloud built on a sparse mixture-of-experts architecture with approximately 1 trillion total parameters.
Qwen Image Max is an image generation model from Qwen for visual creation workflows. It is aimed at illustration, design exploration, and marketing creative.
Qwen Image Plus is an image generation model from Qwen for visual creation workflows. It is aimed at illustration, design exploration, and marketing creative.
Qwen Max is a proprietary frontier model from Alibaba Cloud built on a sparse mixture-of-experts architecture with approximately 1 trillion total parameters.
Qwen MT Flash is a language model from Qwen for chat, writing, analysis, and general-purpose automation. It is tuned for lower latency and cost-sensitive production use.
Qwen MT Lite is a language model from Qwen for chat, writing, analysis, and general-purpose automation. It is tuned for lower latency and cost-sensitive production use.
Qwen MT Plus is a 1 million context hybrid reasoning model with a balanced performance, speed, and cost combination. It accepts text inputs and produces text outputs.
Qwen MT Turbo is a language model from Qwen for chat, writing, analysis, and general-purpose automation. It is tuned for lower latency and cost-sensitive production use.
Qwen Omni Turbo is a speech-to-text model from Qwen for transcription and spoken-language understanding workflows. It is suited to captioning, meeting notes, transcription, and speech analytics pipelines.
Qwen Plus 2025 01 25 is a 131K context model with a balanced performance, speed, and cost combination. It accepts text inputs and produces text outputs.
Qwen Plus 2025 04 28 is a 131K context model with a balanced performance, speed, and cost combination. It accepts text inputs and produces text outputs.
Qwen Plus 2025 07 14 is a 131K context model with a balanced performance, speed, and cost combination. It accepts text inputs and produces text outputs.
Qwen Plus 2025 07 28 is a 1 million context hybrid reasoning model with a balanced performance, speed, and cost combination. It accepts text inputs and produces text outputs.
Qwen Plus 2025 09 11 is a 131K context model with a balanced performance, speed, and cost combination. It accepts text inputs and produces text outputs.
Qwen Plus 2025 12 01 is a 131K context model with a balanced performance, speed, and cost combination. It accepts text inputs and produces text outputs.
Qwen Plus Character is a 1 million context hybrid reasoning model with a balanced performance, speed, and cost combination. It accepts text inputs and produces text outputs.
Qwen Turbo 2024 11 01 is a language model from Qwen for chat, writing, analysis, and general-purpose automation. It is tuned for lower latency and cost-sensitive production use.
Qwen Turbo 2025 04 28 is a language model from Qwen for chat, writing, analysis, and general-purpose automation. It is tuned for lower latency and cost-sensitive production use.
Qwen VL Max 2025 04 08 is a video generation model from Qwen for prompt-driven clip and scene creation. It supports turning still images into animated clips as well as prompt-driven video workflows.
Qwen VL Max 2025 08 13 is a video generation model from Qwen for prompt-driven clip and scene creation. It supports turning still images into animated clips as well as prompt-driven video workflows.
Qwen VL Plus 2025 01 25 is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window.
Qwen VL Plus 2025 05 07 is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window.
Qwen VL Plus 2025 08 15 is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window.
Qwen3 30b is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces.
Qwen3 30b A3b 2507 is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces. It accepts text inputs and produces text outputs.
Qwen3 Coder is a Mixture-of-Experts (MoE) code generation model developed by the Qwen team. It accepts text inputs and produces text outputs.
Qwen3 Coder Flash is Alibaba's fast and cost efficient version of their proprietary Qwen3 Coder Plus. It is tuned for lower latency and cost-sensitive production use.
Qwen3 Coder Plus 2025 07 22 is Alibaba's proprietary version of the Open Source Qwen3 Coder 480B A35B. It accepts text inputs and produces text outputs.
Qwen3 Coder Plus 2025 09 23 is Alibaba's proprietary version of the Open Source Qwen3 Coder 480B A35B. It accepts text inputs and produces text outputs.
Qwen3 Max is an updated release built on the Qwen3 series, offering major improvements in reasoning, instruction following, multilingual support, and long-tail knowledge coverage compared to the January 2025 version.
Qwen3 Max 2025 09 23 is an updated release built on the Qwen3 series, offering major improvements in reasoning, instruction following, multilingual support, and long-tail knowledge coverage compared to the January 2025 version.
Qwen3 Max 2026 01 23 is an updated release built on the Qwen3 series, offering major improvements in reasoning, instruction following, multilingual support, and long-tail knowledge coverage compared to the January 2025 version.
Qwen3 Max Preview is a proprietary frontier model from Alibaba Cloud built on a sparse mixture-of-experts architecture with approximately 1 trillion total parameters.
Qwen3 Next 80b is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces. It accepts text inputs and produces text outputs.
Qwen3 Next 80b A3b is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces. It accepts text inputs and produces text outputs.
Qwen3 Omni 30b A3b is a speech-to-text model from Qwen for transcription and spoken-language understanding workflows. It is suited to captioning, meeting notes, transcription, and speech analytics pipelines.
Qwen3 VL 235b A22b is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window.
Qwen3 VL 30b A3b is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per token. It supports turning still images into animated clips as well as prompt-driven video workflows.
Qwen3 VL 32b is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video.
Qwen3 VL 8b is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window.
Qwen3 VL Flash 2025 10 15 is a fast, efficient language model from Alibaba's Qwen 3.6 series. It supports text, image, and video input with a 1M token context window.
Qwen3 VL Flash 2026 01 22 is a fast, efficient language model from Alibaba's Qwen 3.6 series. It supports text, image, and video input with a 1M token context window.
Qwen3 VL Plus 2025 09 23 is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window.
Qwen3 VL Plus 2025 12 19 is a large-scale multimodal language model from Alibaba. It accepts text, image, and video input and produces text output, with a 1M token context window.
Qwq Plus is a 1 million context hybrid reasoning model with a balanced performance, speed, and cost combination. It accepts text inputs and produces text outputs.
Text Embedding V3 is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. It is aimed at semantic search, retrieval, clustering, and RAG pipelines.
Text Embedding V4 is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. It is aimed at semantic search, retrieval, clustering, and RAG pipelines.
Wan2.2-T2V-A14B is a DigitalOcean-hosted text-to-video model exposed through DigitalOcean Inference.
Code Qwen 1.5 7B is a coding model from Qwen for code generation, edits, and software tasks.
QvQ 72B Preview is a language model from Qwen for chat, writing, analysis, and general-purpose automation. This entry tracks a preview release and capabilities may change before general availability.
Qwen 1.5 0.5B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 1.5 1.8B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 1.5 110B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 1.5 14B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 1.5 32B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 1.5 4B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 1.5 72B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 1.5 7B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 1.5 MoE A2.7B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 1.8B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 14B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 2 0.5B is the latest series of Qwen large language models.
Qwen 2 1.5B is the latest series of Qwen large language models.
Qwen 2 72B Instruct is the latest series of Qwen large language models.
Qwen 2 7B Instruct is the latest series of Qwen large language models.
Qwen 2 Audio 7B is the latest series of Qwen large language models.
Qwen 2 Math 1.5B is the latest series of Qwen large language models.
Qwen 2 Math 72B is the latest series of Qwen large language models.
Qwen 2 Math 7B is the latest series of Qwen large language models.
Qwen 2 Math RM 72B is the latest series of Qwen large language models.
Qwen 2 VL 2B is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Qwen 2 VL 72B is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Qwen 2 VL 7B is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Qwen 2.5 0.5B is the latest series of Qwen large language models.
Qwen 2.5 1.5B is the latest series of Qwen large language models.
Qwen 2.5 14B is the latest series of Qwen large language models.
Qwen 2.5 32B is the latest series of Qwen large language models.
Qwen 2.5 3B is the latest series of Qwen large language models.
Qwen 2.5 72B is the latest series of Qwen large language models.
Qwen 2.5 7B is the latest series of Qwen large language models.
Qwen 2.5 Coder 0.5B is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen).
Qwen 2.5 Coder 1.5B is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen).
Qwen 2.5 Coder 14B is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen).
Qwen 2.5 Coder 32B Instruct is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen).
Qwen 2.5 Coder 3B is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen).
Qwen 2.5 Coder 7B is the latest series of Code-Specific Qwen large language models (formerly known as CodeQwen).
Qwen 2.5 Math 1.5B is the latest series of Qwen large language models.
Qwen 2.5 Math 72B is the latest series of Qwen large language models.
Qwen 2.5 Math 7B is the latest series of Qwen large language models.
Qwen 2.5 Math 7B PRM800K is the latest series of Qwen large language models.
Qwen 2.5 Math PRM 72B is the latest series of Qwen large language models.
Qwen 2.5 Math PRM 7B is the latest series of Qwen large language models.
Qwen 2.5 Math RM 72B is the latest series of Qwen large language models.
Qwen 2.5 Omni 3B is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Qwen 2.5 Omni 7B is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Qwen 2.5 VL 3B Instruct is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Qwen 2.5 VL 7B Instruct is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.
Qwen 3 0.6B is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces.
Qwen 3 1.7B is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces.
Qwen 3 14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue.
Qwen 3 235B A22B is a 235B parameter mixture-of-experts (MoE) model developed by Qwen, activating 22B parameters per forward pass.
Qwen 3 235B A22B Thinking 2507 is a high-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks. It supports up to 262.1K of input context for long documents, codebases, and 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.
Qwen 3 30B A3B Instruct 2507 is a 30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference.
Qwen 3 30B A3B Thinking 2507 is a 30B parameter Mixture-of-Experts reasoning model optimized for complex tasks requiring extended multi-step thinking.
Qwen 3 32B is a dense 32.8B parameter causal language model from the Qwen3 series, optimized for both complex reasoning and efficient dialogue.
Qwen 3 4B is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces.
Qwen 3 4B Instruct 2507 is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces.
Qwen 3 4B SafeRL is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces.
Qwen 3 4B Thinking 2507 is a reasoning-first chat model in the Qwen3-Next line that outputs structured "thinking" traces by default.
Qwen 3 8B is a dense 8.2B parameter causal language model from the Qwen3 series, designed for both reasoning-heavy tasks and efficient dialogue.
Qwen 3 A235 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.
Qwen 3 ASR 0.6B is a speech-to-text model from Qwen for transcription and spoken-language understanding workflows. It is suited to captioning, meeting notes, transcription, and speech analytics pipelines.
Qwen 3 ASR 1.7B is a speech-to-text model from Qwen for transcription and spoken-language understanding workflows. It is suited to captioning, meeting notes, transcription, and speech analytics pipelines.
Qwen 3 Coder 30B A3B Instruct is a 30.5B parameter Mixture-of-Experts (MoE) model with 128 experts (8 active per forward pass), designed for advanced code generation, repository-scale understanding, and agentic tool use.
Qwen3 Coder 480B A35B Instruct is Alibaba's 480B-total-parameter, 35B-active-parameter sparse MoE coding model for repository-scale understanding, tool use, and long-horizon software-engineering agents.
Qwen 3 Coder Next is an open-weight causal language model optimized for coding agents and local development workflows.
Qwen 3 Embedding 0.6B is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. It is aimed at semantic search, retrieval, clustering, and RAG pipelines.
Qwen 3 Embedding 4B is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. It is aimed at semantic search, retrieval, clustering, and RAG pipelines.
Qwen 3 ForcedAligner 0.6B is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces.
Qwen 3 Guard Gen 4B is a safety model from Qwen for moderation, prompt filtering, and policy enforcement. It is suited to input filtering, safety checks, and policy enforcement in LLM applications.
Qwen 3 Guard Stream 0.6B is a safety model from Qwen for moderation, prompt filtering, and policy enforcement. It is suited to input filtering, safety checks, and policy enforcement in LLM applications.
Qwen 3 Guard Stream 4B is a safety model from Qwen for moderation, prompt filtering, and policy enforcement. It is suited to input filtering, safety checks, and policy enforcement in LLM applications.
Qwen 3 Guard Stream 8B is a safety model from Qwen for moderation, prompt filtering, and policy enforcement. It is suited to input filtering, safety checks, and policy enforcement in LLM applications.
Qwen 3 Next 80B A3B Instruct is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without "thinking" traces.
Qwen 3 Next 80B A3B Thinking is a reasoning-first chat model in the Qwen3-Next line that outputs structured "thinking" traces by default.
Qwen 3 Omni 30B A3B Captioner is a multimodal model that unifies strong text generation with visual understanding for images and videos. Its Instruct variant optimizes instruction-following for general multimodal tasks.
Qwen 3 Omni 30B A3B Instruct is a multimodal model that unifies strong text generation with visual understanding for images and videos. Its Instruct variant optimizes instruction-following for general multimodal tasks.
Qwen 3 Omni 30B A3B Thinking is a multimodal model that unifies strong text generation with visual understanding for images and videos. Its Thinking variant enhances reasoning in STEM, math, and complex tasks.
Qwen 3 Omni Flash (2025-09-15) is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video.
Qwen 3 Omni Flash (2025-12-01) is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video.
Qwen 3 Reranker 0.6B is a reranking model from Qwen for improving retrieval quality by rescoring candidate results against a query. Use it to rescore retrieved passages, products, or documents before final ranking or generation.
Qwen 3 Reranker 4B is a reranking model from Qwen for improving retrieval quality by rescoring candidate results against a query. Use it to rescore retrieved passages, products, or documents before final ranking or generation.
Qwen 3 TTS (2025-11-27) is a text-to-speech model from Qwen for turning text into spoken audio. It is suited to narration, assistants, and spoken user interfaces.
Qwen 3 TTS 12Hz 0.6B Base is a text-to-speech model from Qwen for turning text into spoken audio. It is suited to narration, assistants, and spoken user interfaces.
Qwen 3 TTS 12Hz 0.6B CustomVoice is a text-to-speech model from Qwen for turning text into spoken audio. It is suited to narration, assistants, and spoken user interfaces.
Qwen 3 TTS 12Hz 1.7B Base is a text-to-speech model from Qwen for turning text into spoken audio. It is suited to narration, assistants, and spoken user interfaces.
Qwen 3 TTS 12Hz 1.7B VoiceDesign is a text-to-speech model from Qwen for turning text into spoken audio. It is suited to narration, assistants, and spoken user interfaces.
Qwen 3 VL 235B A22B Instruct is an open-weight multimodal model that unifies strong text generation with visual understanding across images and video.
Qwen 3 VL 235B A22B Thinking is a multimodal model that unifies strong text generation with visual understanding across images and video.
Qwen 3 VL 2B Instruct is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video.
Qwen 3 VL 2B Thinking is the reasoning-optimized variant of the Qwen3-VL-8B multimodal model, designed for advanced visual and textual reasoning across complex scenes, documents, and temporal sequences.
Qwen 3 VL 30B A3B Instruct is a multimodal model that unifies strong text generation with visual understanding for images and videos. Its Instruct variant optimizes instruction-following for general multimodal tasks.
Qwen 3 VL 30B A3B Thinking is a multimodal model that unifies strong text generation with visual understanding for images and videos. Its Thinking variant enhances reasoning in STEM, math, and complex tasks.
Qwen 3 VL 32B Instruct is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video.
Qwen 3 VL 32B Thinking is the reasoning-optimized variant of the Qwen3-VL-8B multimodal model, designed for advanced visual and textual reasoning across complex scenes, documents, and temporal sequences.
Qwen 3 VL 4B Instruct is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video.
Qwen 3 VL 4B Thinking is the reasoning-optimized variant of the Qwen3-VL-8B multimodal model, designed for advanced visual and textual reasoning across complex scenes, documents, and temporal sequences.
Qwen 3 VL 8B Instruct is a multimodal vision-language model from the Qwen3-VL series, built for high-fidelity understanding and reasoning across text, images, and video.
Qwen 3 VL 8B Thinking is the reasoning-optimized variant of the Qwen3-VL-8B multimodal model, designed for advanced visual and textual reasoning across complex scenes, documents, and temporal sequences.
Qwen 3 VL Embedding 2B is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. It is aimed at semantic search, retrieval, clustering, and RAG pipelines.
Qwen 3 VL Embedding 8B is the latest proprietary model of the Qwen family, specifically designed for text embedding and ranking tasks. It is aimed at semantic search, retrieval, clustering, and RAG pipelines.
Qwen 3 VL Reranker 2B is a reranking model from Qwen for improving retrieval quality by rescoring candidate results against a query. Use it to rescore retrieved passages, products, or documents before final ranking or generation.
Qwen 3 VL Reranker 8B is a reranking model from Qwen for improving retrieval quality by rescoring candidate results against a query. Use it to rescore retrieved passages, products, or documents before final ranking or generation.
Qwen 72B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen 7B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen Audio is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen Audio Chat is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen Image is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen Image 2512 is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen Image Edit is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen Image Edit 2509 is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen Image Edit 2511 is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen Image Layered is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
Qwen VL is a large-scale multimodal vision-language model designed for high-precision understanding and reasoning across text, images, and video.
QwQ 32B is a reasoning model from Qwen for multi-step analysis, math, and tool-driven tasks.
QwQ 32B Preview is a reasoning model from Qwen for multi-step analysis, math, and tool-driven tasks. This entry tracks a preview release and capabilities may change before general availability.
WorldPM 72B is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
WorldPM 72B HelpSteer2 is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
WorldPM 72B RLHFLow is a language model from Qwen for chat, writing, analysis, and general-purpose automation.
WorldPM 72B UltraFeedback is a language model from Qwen for chat, writing, analysis, and general-purpose automation.