# Qwen (model family)
Qwen is a family of large language models and large multimodal models published by the Qwen Team of Alibaba Group, spanning text, vision, audio, tool use, and agents.[1](https://qwen.readthedocs.io/en/stable/index.html) First released in 2023 and iterated through Qwen2 (2024), Qwen3 (2025), and the Qwen3.5/Qwen3.8 generation (2026), the family is distributed with openly downloadable weights on [Hugging Face](https://www.edgepedia.org/hugging-face) and [ModelScope](https://www.edgepedia.org/modelscope).[2](https://github.com/QwenLM/Qwen) This article covers the model family itself; Alibaba as a company, the Tongyi consumer business, and the Qwen assistant app are separate subjects.
A caveat applies throughout: the published record drawn on here consists almost entirely of Alibaba's own documentation, technical reports, and repositories.
| Fact | Detail |
|---|---|
| Maker | Qwen Team, Alibaba Group[1](https://qwen.readthedocs.io/en/stable/index.html) |
| First release | 2023, original Qwen series (Qwen-1.8B to Qwen-72B)[2](https://github.com/QwenLM/Qwen) |
| Latest generation | Qwen3.8 (2026), described by Alibaba as the first Qwen-Max-class model in open release[4](https://github.com/QwenLM/Qwen3.5?file=Qwen3.5) |
| Size range | 0.5B to 235B total parameters across generations (0.6B–32B dense and up to 235B-A22B MoE in Qwen3)[3](https://arxiv.org/html/2407.10671) • [5](https://qwen.readthedocs.io/en/latest/getting_started/concepts.html) |
| Language coverage | ~30 languages (Qwen2), 119 (Qwen3), 201 languages, and dialects (Qwen3.5)[3](https://arxiv.org/html/2407.10671) • [5](https://qwen.readthedocs.io/en/latest/getting_started/concepts.html) • [4](https://github.com/QwenLM/Qwen3.5?file=Qwen3.5) |
| Licence | Apache 2.0 for repository source code; weights released on Hugging Face and ModelScope[2](https://github.com/QwenLM/Qwen) • [3](https://arxiv.org/html/2407.10671) |
| Derivative models | Over 200,000 as of January 2026, per a weakly sourced encyclopedia entry[6](https://baike.baidu.com/en/item/Qwen/1530291) |
## Release timeline and versions
The original Qwen series appeared in 2023 as base models at 1.8B, 7B, 14B, and 72B parameters, with aligned Qwen-Chat variants on top. Alibaba stated the base models were pretrained on up to 3 trillion tokens of multilingual data with a focus on Chinese and English.[2](https://github.com/QwenLM/Qwen)
Qwen2 followed in July 2024, described in a technical report as four dense models at 0.5B, 1.5B, 7B, and 72B parameters plus a Mixture-of-Experts model with 57 billion total parameters, of which 14 billion are activated per token.[3](https://arxiv.org/html/2407.10671) Mixture-of-Experts (MoE) architectures activate only part of a model's parameters for each token, trading total size for cheaper inference.
Qwen3 shipped in April 2025 with dense models at 0.6B, 1.7B, 4B, 8B, 14B, and 32B parameters and a return of MoE at 30B-A3B and 235B-A22B (30 billion total with 3 billion active, and 235 billion total with 22 billion active).[1](https://qwen.readthedocs.io/en/stable/index.html) • [5](https://qwen.readthedocs.io/en/latest/getting_started/concepts.html) The 2026 generation, Qwen3.5 and the derived Qwen3.8, extended the family to a flagship open release; Alibaba describes Qwen3.8 as the first Qwen-Max-class model brought to open release, targeting coding, professional work, research, and long-horizon agentic tasks.[4](https://github.com/QwenLM/Qwen3.5?file=Qwen3.5)
## Architecture and training as published
Each generation's disclosed training setup grew substantially. Qwen2's models were pre-trained on over 7 trillion tokens of multilingual data and post-trained with supervised fine-tuning followed by direct preference optimization (DPO), aligning the models with human preferences through learning from human feedback.[3](https://arxiv.org/html/2407.10671) The original Qwen's chat models had used SFT and RLHF.[2](https://github.com/QwenLM/Qwen)
Qwen3's documented specifics include a packed pre-training sequence length of 32,768 tokens, extendable to 131,072 tokens where model cards say so, and training on 119 languages and dialects.[5](https://qwen.readthedocs.io/en/latest/getting_started/concepts.html) Its most distinctive published feature is hybrid thinking mode: a single model that can switch between a thinking mode for complex logical reasoning, math, and coding, and a non-thinking instruct mode for general chat, without deploying two separate models.[5](https://qwen.readthedocs.io/en/latest/getting_started/concepts.html)
For Qwen3.5, Alibaba reports early-fusion vision-language training on trillions of multimodal tokens (early fusion means text and images are processed by one model from the start of training), a hybrid architecture combining Gated Delta Networks with sparse Mixture-of-Experts for high-throughput, low-latency inference, and reinforcement learning scaled across million-agent environments, with claimed near-100% multimodal training efficiency relative to text-only training.[4](https://github.com/QwenLM/Qwen3.5?file=Qwen3.5) These are vendor disclosures; the underlying training data composition has not been published in detail.
## Benchmarks: vendor claims only
All quantitative performance figures available for Qwen come from Alibaba. For the original series, the Qwen repository states that Qwen-72B outperforms LLaMA2-70B on all evaluated tasks and beats GPT-3.5 on 7 of 10 tasks, including MMLU, C-Eval, GSM8K, MATH, HumanEval, MBPP, and BBH.[2](https://github.com/QwenLM/Qwen) The Qwen2 technical report gives Qwen2-72B base scores of 84.2 on MMLU, 37.9 on GPQA, 64.6 on [HumanEval](https://www.edgepedia.org/humaneval), 89.5 on GSM8K, and 82.4 on BBH, and Qwen2-72B-Instruct scores of 9.1 on MT-Bench, 48.1 on [Arena-Hard](https://www.edgepedia.org/arena-hard), and 35.7 on [LiveCodeBench](https://www.edgepedia.org/livecodebench).[3](https://arxiv.org/html/2407.10671) Alibaba also claims leading performance among open-source models on complex agent-based tasks for Qwen3.[1](https://qwen.readthedocs.io/en/stable/index.html)
Readers should treat them as the vendor's own measurements, produced under the vendor's chosen harnesses and task selections.
## Licensing and availability
The QwenLM/Qwen repository source code is licensed under Apache 2.0, and Qwen2 weights were released openly on Hugging Face and ModelScope with code and fine-tuning resources on GitHub.[2](https://github.com/QwenLM/Qwen) • [3](https://arxiv.org/html/2407.10671)
## Adoption and what changed in 2025–2026
The main documented adoption signal is a derivative-model count: as of January 2026, the number of Qwen derivative models exceeded 200,000, which the source describes as a first among open-source large model families.[6](https://baike.baidu.com/en/item/Qwen/1530291) This figure comes from a user-edited encyclopedia entry and should be treated as weakly sourced; no independent download statistics or enterprise deployment data are available here.
The 2025–2026 changes visible in the vendor record are threefold: Qwen3's hybrid thinking and non-thinking modes within one model, simplifying deployment;[5](https://qwen.readthedocs.io/en/latest/getting_started/concepts.html) Qwen3.5's early-fusion multimodal training and Gated Delta Network plus sparse MoE architecture;[4](https://github.com/QwenLM/Qwen3.5?file=Qwen3.5) and the open release of a Qwen-Max-class flagship in Qwen3.8, which previously existed only as a closed API-tier model.[4](https://github.com/QwenLM/Qwen3.5?file=Qwen3.5)
## References
1. Qwen official documentation (readthedocs, stable) — https://qwen.readthedocs.io/en/stable/index.html
2. QwenLM/Qwen — official GitHub repository (original Qwen release, 2023) — https://github.com/QwenLM/Qwen
3. Qwen2 Technical Report (arXiv, Alibaba Qwen Team, July 2024) — https://arxiv.org/html/2407.10671
4. QwenLM/Qwen3.5 — official repository (Qwen3.5 / Qwen3.8, 2026) — https://github.com/QwenLM/Qwen3.5?file=Qwen3.5
5. Qwen documentation — Key Concepts (Qwen3) — https://qwen.readthedocs.io/en/latest/getting_started/concepts.html
6. Qwen (Alibaba Tongyi Lab series of models) — Baidu Baike — https://baike.baidu.com/en/item/Qwen/1530291
---
*Topic: Encyclopedia › Technology and the built world › Computing and digital systems › Modern AI: foundation models, generative AI, and the AI industry › Model families and named models › Large language model families › Open-weight model families*
*Initially written Sep 17, 2026 · Reviewed: Sep 19, 2026 · Edited: Sep 19, 2026; Oct 11, 2026; Oct 15, 2026 · Last review: Sep 19, 2026*
*Copyright 2026 EdgeChat AI, a subsidiary of Biostate AI.*
License: Edgepedia Community License 1.0, https://www.edgepedia.org/edgepedia/license