한국어EN
Alibaba

The Qwen family

Alibaba's Qwen models, released as open weights across a wide range of sizes.

Only versions we have a record for. Missing ones appear as records are added.

13versions2023–2026Latest · Qwen3.8-Max

Made by AIThe family notes were written by AI (Claude); each version links to its record, where the source is. Published after automated checks; the owner doesn't read every item.How it's made

At a glance

The Qwen family — 13 versionsEach Qwen version from 2023 to 2026 as a dot, in release order. The list below has the same information.Main lineQwen-7B · Aug 3, 2023Qwen2 · Jun 7, 2024Qwen2.5 · Sep 19, 2024QwQ-32B-Preview · Nov 28, 2024QwQ-32B · Mar 6, 2025Qwen3 · Apr 29, 2025Qwen3-Coder · Jul 22, 2025Qwen3.5-397B-A17B · Feb 2026Qwen3.5-Omni · Apr 17, 2026Qwen3.6-27B · Apr 22, 2026Qwen3.8-2.4T-A95B · Aug 2026Qwen3.8-27B · Aug 2026Qwen3.8-Max · Aug 4, 2026
Where labels would overlap, only the dot is drawn — point at it to see the name; every version is in the list below.

13 versions

In release order. Each date comes from the record that documents it — open a name to see the record.

  1. Qwen-7B

    The first open-weight Qwen

    TextOpen weights

  2. Qwen2

    Five sizes, four of them under Apache 2.0

    TextOpen weights

  3. Qwen2.5

    Seven sizes from 0.5B to 72B

    TextOpen weights

  4. QwQ-32B-Preview

    An open reasoning-model preview that followed o1

    ReasoningOpen weights

  5. QwQ-32B

    A 32B reasoning model trained with reinforcement learning

    ReasoningOpen weights

  6. Qwen3

    Every size from 0.6B to 235B under Apache 2.0

    Open weights

  7. Qwen3-Coder

    A 480B mixture-of-experts code model

    CodeOpen weights

  8. Qwen3.5-397B-A17B

    Large MoE with a vision encoder, under Apache 2.0

    Open weightsMultimodal

  9. Qwen3.5-Omni

    Technical report for an omni-modal Thinker-Talker model

    Multimodal

  10. Qwen3.6-27B

    A dense 27B model released under Apache 2.0

    Open weights

  11. Qwen3.8-2.4T-A95B

    Top open-weight model, under a custom license, not Apache

    Open weights

  12. Qwen3.8-27B

    27B dense model with a vision encoder, still Apache 2.0

    Open weightsMultimodal

  13. Qwen3.8-Max

    Promised open weights, but none were out at launch

    LatestText