DeepSeek V3.2 Exp

DeepSeek · Mid

Compare →

DeepSeek-V3.2-Exp is an experimental large language model released by DeepSeek as an intermediate step between V3.1 and future architectures. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...

Input / 1M

$0.27

Output / 1M

$0.41

Cached input / 1M

Context window

163K

Where it sits

84th-cheapest mid model

by blended $/Mtok among 262 listed mid models

47% below the mid median

blended $/Mtok across 262 mid models

Output costs 1.5× input

$0.27 in / $0.41 out per 1M

Held flat since launch (Jun 2026)

no blended price change recorded

Computed live from current prices and this model's history — not hand-written, so it stays accurate as prices move.

Price history

Only one price on record so far — the history chart appears once a price changes.

Snapshots

Effective Input Output Cached in Note Source
11 Jun 2026 $0.27 $0.41 Imported from OpenRouter openrouter.ai

More from DeepSeek