Qwen Embedding · Reranking · Qwen3 Reranker

Qwen3 Reranker 8B: Specs, GPU requirements, hosting & deployment

Qwen3 Reranker 8B is a current Reranking release in the Qwen Embedding family, certified from its immutable official source revision.

Calculate recommendation Customise a model Enterprise Model Transformation Compare this model

Technical deployment reference illustration for Qwen3 Reranker 8B — Reference deployment flow; exact architecture facts are listed separately.

Total parameters8.0B

Active parametersN/A / Not directly comparable

Context window32,768 tokens

Release statusCurrent · Current

ArchitectureReranking transformer

ModalitiesText

LicenseApache License 2.0

Deployment stateDeployable with certified configuration

Commercial stateCommercial Use Allowed

Release date2025-05-29

Architecture, strengths and use cases

Qwen3 Reranker 8B is an official self-hostable Qwen Embedding and reranking release. It uses Reranking transformer with 8B. Exact checkpoint, license, runtime, and hardware evidence is tied to the verified upstream revision.

Qwen3 Reranker 8B uses Reranking transformer with 8B. Pipeline and configuration fields are reported only where the official source publishes them.

Verified Reranking transformer architecture

Official checkpoint and pipeline evidence

Official BF16, Unknown checkpoint coverage

Self-hosted Embedding and reranking workloads

Controlled model evaluation

Production deployment planning

Certified checkpoints and quantizations

Only exact, source-certified variants attached to this release are shown.

Qwen3-Reranker-8B

Qwen/Qwen3-Reranker-8B

Official · BF16 · safetensors

Runtime certified: Transformers

Precision / quantization	Format	Bits	Publisher trust	Deployment state
BF16	safetensors	16	Official	Runtime certified

Recommended hardware

Use only the listed runtime configurations with the exact checkpoint precision. Hardware sizing considers resident pipeline memory, runtime overhead, input and output dimensions, batch size, and safety headroom.

Estimated weight memory15.35 GB

Runtime overhead7.48 GB

Total with safety margin28.31 GB

Recommended best value

1× NVIDIA A100 SXM 80GB

80 GB total VRAM · nvlink

51.69 GB headroom · Available

USD 915 / month

Lowest-cost viable

2× NVIDIA L4 24GB

48 GB total VRAM · pcie

19.69 GB headroom · Available

USD 500 / month

Highest performance

1× NVIDIA H200 SXM 141GB

141 GB total VRAM · nvswitch

112.69 GB headroom · Limited capacity

USD 2,345 / month

License and self-hosting considerations

The exact pipeline and selected precision must fit on the chosen GPUs. Requirements vary with resolution, duration, batch size, modalities, and runtime; language-model KV-cache rules are not applied.

Qwen3 Reranker 8B is published under Apache License 2.0. The release-scoped official license link and verification date are retained with the catalogue record.

Commercial Use Allowed. The official license permits commercial use; its obligations still apply.

Alternatives and sibling variants

千问3-Reranker-8B

Release timeline

qwen/qwen-embedding/qwen3-reranker-8b

3 Sept 2026

Source details

Official source

Source revision: 01877eea9646ea6f04f16d6f79648f3c512bca5fb1b2b760649934b08e7c496b

Frequently asked questions

How much VRAM does Qwen3 Reranker 8B require?

The estimate uses the certified checkpoint, precision, runtime, workload dimensions, batch size, and domain-specific overhead.

Which runtimes support Qwen3 Reranker 8B?

Certified upstream runtime evidence is available for transformers.

Can Qwen3 Reranker 8B be used commercially?

The stored commercial classification is Commercial Use Allowed. Consult the official Apache License 2.0 terms for the final obligations.

Official references

Last verified: 2026-08-26