DeepSeek V3.2 Exp

deepseek/deepseek-v3.2-exp

DeepSeek V3.2 Exp is a text‑only language model released by DeepSeek on 2025‑01‑01. It supports up to 163,840 tokens of context, offers tool use and reasoning capabilities, and is provided with open weights. Pricing is $0.27 per million input tokens and $0.41 per million output tokens.

  • tool use
  • structured outputs
  • reasoning
  • open-weight

At a glance

providerdeepseek
releasedJan 1, 2025
context164K
input / M$0.27

Specifications

What the API gives you

Context window
164K tokens
Max output
66K tokens
Input price
$0.27 / 1M tokens
Output price
$0.41 / 1M tokens
Cache read
Modalities
text → text
Knowledge cutoff
Jul 31, 2025
Released
Jan 1, 2025

Recommendation coverage

Where this model appears

Not currently included in a published recommendation. This means the available task-specific evidence did not support a ranking yet—not that the model was reviewed and rejected.

Measured evidence

Public benchmark record

LMArena Text
1423 Elo2026-07-16
LMArena WebDev
1288 Elo2026-07-16
LMArena WebDev
1288 Elo2026-07-20

LMArena data licensed CC-BY-4.0. Aider (Apache-2.0) and SWE-bench (MIT) are open.

Open-weight model on Hugging Face.

Catalog data via OpenRouter and models.dev. Last checked within the past 6 hours.