Qwen3.8-Flash-Next logo

Qwen3.8-Flash-Next

The open-weight preview of Qwen4

▲ 107 Product Hunt votes Launched Aug 27, 2026
Visit Qwen3.8-Flash-Next →

The open-weight preview of Qwen4

Qwen3.8-Flash-Next is a 125B multimodal MoE with only 6B active parameters and a new architecture built around QSA, Gated Residual, N-gram embeddings, and Muon. Its open weights give an early look at the architecture Qwen is building toward Qwen4.

Qwen3.8-Flash-Next launched on Product Hunt on 2026-08-27 and has 98 upvotes and 1 comments so far.

Pricing details for this tool haven’t been verified yet — check the tool’s website for current plans.

Categories: Coding AI
View on Product Hunt · Listing last verified: Aug 28, 2026

More Coding AI tools

ShipHQ Moxie Sayonic Earlyn