Qwen3.8-Flash-Next
Official · BF16 · safetensors
Runtime certified: Transformers vLLM SGLang Tokenspeed
| Precision / quantization | Format | Bits | Publisher trust | Deployment state |
|---|---|---|---|---|
| BF16 | safetensors | 16 | Official | Runtime certified |
Qwen · Fast multimodal · Qwen3.8 Flash Next
Qwen3.8-Flash-Next: Specs, GPU Requirements and License
Qwen3.8-Flash-Next is a current Fast Multimodal release in the Qwen family, certified from its immutable official source revision.
Separate Commercial License Required
Compare this model
Total parameters180.0B
Active parameters6.0B
Context window262,144 tokens
Release statusCurrent · Current
ArchitectureQwen4ExpForConditionalGeneration
ModalitiesText, Image, Video
LicenseQwen Community License 1.0
Deployment stateDeployable with certified configuration
Commercial stateSeparate Commercial License Required
Release date2026-08-24
Architecture, strengths and use cases
Qwen3.8-Flash-Next is an official self-hostable Qwen release in the Qwen Fast Multimodal track. It uses Qwen4ExpForConditionalGeneration, carries 180B / 6B MoE, supports a verified context window of 262,144 tokens, and exposes only source-certified checkpoints and runtime combinations. Its release, license and deployment facts are tied to the official repository revision rather than inferred from the family name.
Qwen3.8-Flash-Next uses Qwen4ExpForConditionalGeneration with 180B / 6B MoE. The stored architecture profile records 48 layers, hidden width 2560, 24 attention heads and a 262,144 token context ceiling where the official configuration publishes those fields.
Verified Qwen4ExpForConditionalGeneration architecture
262,144 token certified context
Official BF16, Unknown checkpoint coverage
Private LLM inference
Agentic workflows
Long-context document processing
Vision-language analysis
Certified checkpoints and quantizations
Only exact, source-certified variants attached to this release are shown.
Recommended hardware
Use only the listed transformers, vllm, sglang, tokenspeed runtime configuration with the exact checkpoint precision. The recommendation engine checks weight memory, KV cache, runtime overhead, safety headroom, GPU architecture and topology before ranking rentable and reference hardware.
Separate Commercial License Required
A separate commercial license or authorization is required. No commercial hosting offer is shown.
License and self-hosting considerations
The full 180B resident weight set must fit across the selected GPUs. Context and concurrency add KV-cache memory; MoE active parameters do not replace resident-weight memory.
Qwen3.8-Flash-Next is published under Qwen Community License 1.0. The release-scoped official license link and verification date are retained with the catalogue record.
Separate Commercial License Required. A separate commercial license or authorization is required. No commercial hosting offer is shown.
Alternatives and sibling variants
千问3.8-Flash-Next, Qwen3.8-Flash-Next-FP8
Release timeline
qwen/qwen/qwen3-8-flash-next
Source details
Source revision: 24b6d65850ab647254a59896ca97b9294561271ff81fa011700f4fe01c4adbe2
Frequently asked questions
How much VRAM does Qwen3.8-Flash-Next require?
The exact estimate depends on checkpoint precision, context and concurrency; the page recommendation uses the certified memory engine and compatible runtime configurations.
Which runtimes support Qwen3.8-Flash-Next?
Certified upstream runtime evidence is available for transformers, vllm, sglang, tokenspeed.
Can Qwen3.8-Flash-Next be used commercially?
The stored commercial classification is Separate Commercial License Required. Consult the official Qwen Community License 1.0 terms for the final obligations.
Official references
Last verified: 2026-08-26