DeepSeek V3.2
Hybrid-reasoning DeepSeek model with thinking and non-thinking modes, sparse attention, and tool-use
Context
163,840
Max output
81,920
Input / 1M
$0.62
Output / 1M
$1.85
Providers
1
01Profile
Profile
Released
2025-12-01
Last updated
2026-02-06
Knowledge cutoff
2024-07
Weights
Open
02Capabilities
Capabilities
Input
text
Output
text
Reasoning
Yes
Tool calling
Yes
Structured output
Yes
Attachments
No
Cache read
—
Cache write
—
03Pricing
Pricing across 1 providers
Per 1M tokens, USD. Sorted by input price. First-party row highlighted.
04History
Change history
1+ recent events since 2025-12-01
2026-02-15Addedamazon-bedrockoffering→ new