Qwen3.5 397B-A17B FP8
Large open Qwen multimodal MoE for visual agents and long technical tasks
Context
200,000
Max output
65,536
Input / 1M
$0.99
Output / 1M
$4.46
Providers
1
01Profile
Profile
Released
2026-02-15
Last updated
2026-08-01
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 1 providers
Per 1M tokens, USD. Sorted by input price. First-party row highlighted.
04History
Change history
7+ recent events since 2026-02-15
2026-08-27Removedneuralwattofferingremoved→ new
2026-08-04Addedinfomaniakoffering→ new
2026-05-16ContextneuralwattContext window262,144→ 262,128
2026-05-16ContextneuralwattMax output262,144→ 262,128
2026-04-29Price upneuralwattinput / 1M$0.35→ $0.69
2026-04-29Price upneuralwattoutput / 1M$1.75→ $4.14
2026-04-27Addedneuralwattoffering→ new