Alibaba Open-Sources Qwen3.8-Flash-Next, an Early Preview of the Qwen4 Architecture

On August 26, Alibaba's Qwen team released and open-sourced the multimodal MoE model Qwen3.8-Flash-Next on Hugging Face and ModelScope, positioning it as an early preview of the Qwen4 architecture, accompanied by a technical report and detailed evaluations. The model has 125B main parameters plus roughly 51B N-gram embedding parameters, activating only 6B per token; officials and media claim training costs dropped to about 1/9 of the previous generation, with Day-0 support from vLLM, SGLang, and NVIDIA NeMo. Community reports say it beats DeepSeek V4 Flash with about half the parameters and surpasses Claude Opus 4.6, topping ClawEval and multilingual SWE-Bench—making it the most-watched open-source release of the day, with coverage from QbitAI and Zhidongxi spreading the news from August 27.

Confirmed

• Official accounts (@AlibabaQwen, @Qwen) confirmed the multimodal MoE architecture: 125B total main parameters, 6B activated per token, with an extreme cost-efficiency pitch (media citing halved pricing and 1/9 training cost).
• Architecture upgrades include GDN+QSA hybrid attention, GatedResidual multi-branch residuals, and N-gram Embedding capacity expansion (51B extra N-gram embedding parameters); community posts relayed details of 36 Gated DeltaNet layers plus 12 Qwen Sparse Attention layers, with 23.5GB memory savings claimed.
• The model supports image-text-to-text input, safetensors format and API compatibility, and an FP8 open-weight release on Hugging Face.
• Pricing and context: Qwen3.8-Flash is live on QwenCloud with native 262K-token context extensible to 1M tokens. Domestic API pricing per million tokens: ¥0.8 input, ¥2.7 output, ¥0.1 cache hit (flat rate per @AlchainHust); overseas: $0.15 input, $0.47 output, $0.016 cache hit per @MaziyarPanahi.
• Day-0 ecosystem support: vLLM, SGLang, and NVIDIA NeMo (NeMo AutoModel and NeMo RL for domain fine-tuning).
• @jiqizhixin confirmed availability on Hugging Face and ModelScope with a technical report and evaluations.
• Benchmarks (community-relayed): Agents' Last Exam score 51.2, Pass@1 24.3%; ranked first on ClawEval and multilingual SWE-Bench, second on DeepSWE, third on SWE-Bench Pro; @bindureddy later claimed it beats GLM 5.3 Flash and DeepSeek Flash, calling it the best Flash model on the market.
• @sammcj opened a discussion thread covering quantization, fine-tuning, chat templates, and inference server support.

Unconfirmed

• @ariG23498 cited 180B parameters, inconsistent with the official 125B + 51B breakdown; treat the official figure as authoritative.
• All comparative claims (beating Claude Opus, DeepSeek V4 Flash, GLM 5.3 Flash) and rankings come from community relay posts; layer-structure and memory-saving details remain unverified against the official report. Superlatives such as "lowest cache-hit price globally" are author opinions; @bindureddy's $0.16 input price differs slightly from @MaziyarPanahi's $0.15.

Why it matters

• As the first preview of the Qwen4 architecture, its ultra-sparse MoE design (6B/125B activation) and new components (GDN+QSA, GatedResidual, N-gram Embedding) signal Alibaba's next-generation direction, with N-gram Embedding echoing DeepSeek's Engram idea—an open-source architecture relay.
• Radically lower costs plus open weights and rapid ecosystem support could directly challenge premium closed models and competing Flash-tier models on price-performance, provided the benchmark claims hold up.

2026-08-26 ~ 2026-08-28 · 26 related posts

Full story(7 episodes)→

• Episode 1: Alibaba's Qwen Team Announces Qwen3.8-Flash-Next as a Preview of Qwen4 Architecture(2026-08-25, 13 posts)
• Episode 2: Alibaba Open-Sources Qwen3.8-Flash-Next, an Early Preview of the Qwen4 Architecture(2026-08-26, 26 posts)
• Episode 3: Qwen3.8-Flash ships with local deployment and claims SWE-bench Pro win over Claude(2026-08-26, 5 posts)
• Episode 4: Qwen3.8-Flash-Next FP8 Quantized Model Released on Hugging Face(2026-08-27, 2 posts)
• Episode 5: Qwen 3.8-Next Released with Detailed Technical Report(2026-08-27, 2 posts)
• Episode 6: Uncensored Qwen3.8-Flash-Next Weights and GGUF Released(2026-08-27, 2 posts)
• Episode 7: Alibaba Open-Sources Qwen3.8-Flash: 125B MoE with 6B Active Params, 20x Cheaper(2026-08-28, 7 posts)

• Qwen 3.8 Flash Next Release: Megathread for Quants, Fine-Tunes, and Benchmarks — sammcj · 2026-08-26
• [source] Alibaba Unveils Qwen3.8-Flash Preview: MoE Model with Next-Gen Architecture — Alibaba_Qwen · 2026-08-26
• Alibaba Qwen Releases Qwen3.8-Flash: 6B Active Params, 9x Cost Reduction — 千问大模型 · 2026-08-26
• Qwen 3.8 Flash-Next: 6B-Active Open Model Beats Claude Opus 4.6 Max — kimmonismus · 2026-08-26
• Alibaba Releases 180B Qwen3.8-Flash-Next Multimodal Model — ariG23498 · 2026-08-26
• [source] Alibaba Releases Multimodal Model Qwen3.8-Flash-Next on Hugging Face — Qwen · 2026-08-26
• Qwen3.8-Flash-Next architecture: 125B params, sparse MoE, 262K context — Alibaba_Qwen · 2026-08-26
• Alibaba Releases Qwen3.8-Flash: Insane Price-Performance, Hints at Qwen4 Arch — jiqizhixin · 2026-08-26
• Qwen 3.8 Flash Next Released: Beats DS V4 Flash with Half the Parameters — elemental-mind · 2026-08-26
• Alibaba's Qwen3.8 with 51B N-gram embeddings now available on SGLang — Alibaba_Qwen · 2026-08-26
• Alibaba releases Qwen3.8-Flash-Next: 90% training cost cut via new architecture — AI寒武纪 · 2026-08-26
• Alibaba releases Qwen3.8-Flash-Next: activates only 6B parameters per token with 50% price cut — 智东西 · 2026-08-27
• Qwen3.8-Flash-Next Launches with Cost-Efficiency Focus — petrusenko_max · 2026-08-27
• Qwen3.8-Flash-Next: GDN+QSA Hybrid Attention Cuts VRAM by 23.5GB — ying11231 · 2026-08-27
• Qwen3.8-Flash-Next Hits 51.2 on Agent Benchmark — qinzytech · 2026-08-27
• Alibaba Open Sources Qwen3.8-Flash-Next with DeepSeek-Inspired Architecture — 量子位 · 2026-08-27
• NVIDIA Provides Day-0 Support for Alibaba's Qwen3.8-Flash-Next with NeMo — Alibaba_Qwen · 2026-08-27
• NVIDIA NeMo announces day-0 support for finetuning Alibaba Qwen — NVIDIAAI · 2026-08-27
• Qwen3.8-Flash-Next Released, Tops ClawEval and Multilingual SWE-Bench — jyangballin · 2026-08-27
• [source] Qwen3.8-Flash launches on QwenCloud with 1M context and aggressive pricing — Alibaba_Qwen · 2026-08-27
• Qwen3.8-Flash cache hits cost just $0.016 per million tokens — MaziyarPanahi · 2026-08-27
• Qwen3.8-Flash slashes prices: flat 2.7 RMB output, cheapest cache hits yet — AlchainHust · 2026-08-27
• Zhipu & Alibaba release new Flash models, redefining cost-performance — APPSO · 2026-08-28
• Qwen Flash outperforms DeepSeek and GLM 5.3 Flash — bindureddy · 2026-08-28

2 near-duplicate retellings: SimplyAnnisa · bindureddy