TellWorldAI World Archive
Ask中文
News
Models
Organisations
People
Papers & concepts
Products & agents
Chips & robots
History
Topics
Ask中文

Home›Models

Model

Qwen2.5-Max

Qwen2.5-Max is Alibaba's large mixture-of-experts model, announced in January 2025. It was offered through Qwen Chat and the Alibaba Cloud API, and its weights were not released.

Status AvailableModality TextOpen weights No Era Generative AI for everyone

License
Proprietary
Access
API, App
Release date
Developer
Alibaba Cloud

What it is

Qwen2.5-Max is a large language model from Alibaba's Qwen team. It uses a mixture-of-experts (MoE) design, in which the network is split into many expert sub-networks and only some of them run for each token. This lets a very large model keep its per-token computation lower than its total size would suggest.

Origins

The Qwen team announced the model in a blog post dated 28 January 2025. Wikipedia gives the launch date as 29 January 2025, one day later.[1][2]

Training used a pre-training corpus of more than 20 trillion tokens. After pre-training, the model was tuned with supervised fine-tuning and reinforcement learning from human feedback (RLHF).[1]

Main features

The Qwen team reported that the instruct version outperformed DeepSeek V3 on Arena-Hard, LiveBench, LiveCodeBench and GPQA-Diamond, and was competitive on MMLU-Pro.[1] The base model was reported to show advantages over DeepSeek V3, Llama-3.1-405B and Qwen2.5-72B on most benchmarks.[1] The team also compared the model with GPT-4o and Claude 3.5 Sonnet.[1]

Access was through two channels. The model was served in Qwen Chat and through the Alibaba Cloud API under the name qwen-max-2025-01-25. The API follows the OpenAI API format, so programs written for that format can call it with minimal changes.[1]

Unlike the open-weight Qwen models, Qwen2.5-Max's weights were not released.[1]

Two training terms in the description are worth explaining. Supervised fine-tuning trains the model on examples of good questions and answers. Reinforcement learning from human feedback (RLHF) uses human judgements of which answers are preferable to steer the model toward those answers. Together they shape a pre-trained model into an assistant. The weights are the learned numbers that determine how a model behaves. Because they were not released, the model can be used only through the provider's service.[1]

Place in development

Qwen2.5-Max was Alibaba's proprietary flagship. In September 2025 that role passed to Qwen3-Max.[2]

CapabilitiesLarge language models

TechniquesFoundation models

Ecosystem & regionChinaCloud

Version history

14
  1. 2023
    • Aug 3, 2023 Qwen (Tongyi Qianwen, 2023)

      Qwen, known in Chinese as Tongyi Qianwen (通义千问), is Alibaba's family of large language models. Its first open-weight model, Qwen-7B, was released on 3 August 2023, and the family has grown through many later versions.

  2. 2024
    • Feb 5, 2024 Qwen1.5

      Qwen1.5 is the Qwen family's open model release of February 2024, with base and chat models from 0.5B to 72B parameters, later extended to 110B, and a mixture-of-experts model.

    • Jun 7, 2024 Qwen2

      Qwen2 is the Qwen family's June 2024 generation, with five sizes from 0.5B to 72B parameters, context lengths up to 128K tokens, and most sizes released under Apache 2.0.

    • Sep 19, 2024 Qwen2.5

      Qwen2.5 is the Qwen family's September 2024 generation. Its launch included 13 open-weight models, with up to 18 trillion pretraining tokens, 128K context and most models under Apache 2.0.

    • Nov 28, 2024 QwQ-32B-Preview

      QwQ-32B-Preview is an experimental reasoning research model released by the Qwen team on 28 November 2024, with 32B parameters, a 32K-token context and open Apache 2.0 weights. It scored strongly on maths benchmarks but the team documented clear failure modes.

  3. 2025
    • Jan 28, 2025 Qwen2.5-Max · This page

      Qwen2.5-Max is Alibaba's large mixture-of-experts model, announced in January 2025. It was offered through Qwen Chat and the Alibaba Cloud API, and its weights were not released.

    • Mar 6, 2025 QwQ-32B

      A 32-billion-parameter open-weight reasoning model released by the Qwen team on 6 March 2025, trained mainly through scaled reinforcement learning and reported to perform comparably to the much larger DeepSeek-R1.

    • Apr 29, 2025 Qwen3

      Qwen3 is a family of large language models from Alibaba, released on 29 April 2025 under Apache 2.0. Its main feature is hybrid thinking, which lets one model reason step by step or answer directly.

    • Jul 21, 2025 Qwen3-2507

      Qwen3-2507 is a staged 2025 refresh of the open-weight Qwen3 models that split thinking and non-thinking into separate checkpoints, raised native context to 262,144 tokens, and scored far above the original non-thinking mode on published benchmarks.

    • Sep 2025 Qwen3-Max

      Qwen3-Max is Alibaba's proprietary trillion-parameter language model, released in September 2025 and served through Alibaba's API. It is the Max-tier flagship that followed Qwen2.5-Max.

    • Sep 2025 Qwen3-Next

      Qwen3-Next is an open-weight model from Alibaba's Qwen team, released in September 2025 under Apache 2.0. It has 80B total parameters, only 3B active per token, and a hybrid attention design later used in Qwen3.5.

  4. 2026
    • Feb 16, 2026 Qwen3.5

      Qwen3.5 is a natively multimodal model family from Alibaba released in February 2026. Its open flagship, Qwen3.5-397B-A17B, has 397B total parameters and 17B active per token, under Apache 2.0.

    • Apr 2, 2026 Qwen3.6

      A generation of Alibaba's Qwen models released in April 2026, spanning proprietary Plus and Max versions aimed at agentic coding and open-weight models such as Qwen3.6-35B-A3B, a 35B-parameter mixture-of-experts with 3B active parameters.

    • Aug 3, 2026 Qwen3.8

      Qwen3.8 is Alibaba's 2026 flagship line. Its open 2.4-trillion-parameter model, Qwen3.8-2.4T-A95B, released in August 2026, has 95B active parameters and a custom license.

20251

  1. ReleaseAlibaba unveils Qwen2.5-Max, claiming wins over DeepSeek V3

    On 28 January 2025 Alibaba unveiled Qwen2.5-Max and said its instruct model beat DeepSeek V3 on several benchmarks. The model's weights were not released, and access went through Qwen Chat and the Alibaba Cloud API.

Developed by1

Alibaba Cloud

Successor of1

Qwen2.5

Succeeded by1

Qwen3-Max

Powers1

Qwen app (Tongyi)

Uses2

Reinforcement learning from human feedback (RLHF)Mixture of experts (MoE)

Notes and sources

  1. Qwen2.5-Max: Exploring the Intelligence of Large-scale MoE Modelhttps://qwenlm.github.io/blog/qwen2.5-max/
  2. Qwenhttps://en.wikipedia.org/wiki/Qwen

Cite this page: Qwen2.5-Max. TellWorld, https://www.tellworld.net/models/qwen2-5-max. Updated 2026-11-08