Qwen3.8 Flash Next
Open-weight experimental preview of the Qwen4 architecture: hybrid-attention MoE (125B total, 6B active) with vision encoder for coding, agent tasks, and image and video understanding
Context
262,144
Max output
262,144
Input / 1M
$0.15
Output / 1M
$0.47
Providers
3
01Profile
Profile
Released
2026-08-27
Last updated
2026-08-27
Knowledge cutoff
—
Weights
Open
02Capabilities
Capabilities
Input
text · image
Output
text
Reasoning
Yes
Tool calling
Yes
Structured output
Yes
Attachments
Yes
Cache read
—
Cache write
—
03Pricing
Pricing across 4 providers
Per 1M tokens, USD. Sorted by input price. First-party row highlighted.
04History
Change history
6+ recent events since 2026-08-27
2026-09-14Removedvercelofferingremoved→ new
2026-09-02Addedrequestyoffering→ new
2026-09-02Addedrequestyoffering→ new
2026-08-31Addedverceloffering→ new
2026-08-28Addedamdoffering→ new
2026-08-28Addedcortecsoffering→ new