Skip to content
AAA.win

AI model family

Qwen3 model family

Qwen's open-weight Qwen3 family, spanning dense and mixture-of-experts checkpoints with instruct and thinking variants.

Editorially reviewedVerified 2026-08-112 official or primary sources

What the family record covers

The official Qwen3 repository documents multiple dense and mixture-of-experts sizes, plus instruct and thinking variants. It also documents local and production serving paths across several inference frameworks.

A Qwen3 record is therefore a family map, not one performance profile. Evaluation must name the checkpoint, revision, quantization, inference framework, context settings, prompt mode, and hardware or hosted endpoint.

Documented access paths

  • Official Qwen model repositories and named model hubs
  • Self-managed inference through supported open serving frameworks
  • Provider-hosted access is a separate deployment whose live catalog must be checked

Capabilities to verify

  • Multilingual instruction following in the languages the workflow actually serves
  • Thinking versus instruct behavior on the same acceptance set
  • Tool-call formatting under the chosen serving framework
  • Quality and resource impact of checkpoint size and quantization

Decision questions

  1. Which Qwen3 checkpoint and revision are within the deployment budget?
  2. Should thinking and instruct workloads use separate pinned variants?
  3. Which serving framework and prompt template will be treated as production configuration?

Evaluation cautions

  • Qwen3 family members are not interchangeable across sizes and modes.
  • Third-party model labels may not map cleanly to the official checkpoint name.
  • Self-hosting shifts availability, security, and upgrade responsibility to the operator.

Related decisions

Move from a family name to the product, provider, workflow, comparison, or protocol that defines the real choice.

Sources checked

Open the original pages before relying on a time-sensitive product decision.

  1. Qwen3 official repositoryQwen Team
  2. Qwen3 conceptsQwen Team
Version · v4.3.4-indexnow-root-proof

Latest releases

IndexNow root-proof request compatibility

After v4.3.3, the exact root-level {key}.txt proof returned HTTP 200, but a full request carrying keyLocation still returned HTTP 403; a minimal homepage request with the same production key and no keyLocation returned HTTP 202. This release omits that field, while the automatic full run and idempotent rerun remain deployment checks.

IndexNow key-proof compatibility

Changed IndexNow verification to the official root-level {key}.txt convention after the first production notification returned HTTP 403; revalidation remains pending, while the website, sitemaps, and Bing sitemap processing are unaffected.

Bing sitemap discovery and IndexNow change notifications

Prepared canonical sitemap discovery for Bing and added automatic IndexNow change notifications while keeping segmented sitemaps authoritative and making no claim that a notified URL has been crawled or indexed.

View all releases