/ Models

Language models

DeepSeek: DeepSeek V3.2 Exp

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...

Maker
Deepseek
Context window
164K
Max output
66K
Input / 1M
$0.27
Output / 1M
$0.41
1M in + 1M out
$0.68

Everything else

Max output66K
Input / 1M$0.27
Output / 1M$0.41
Cache read / 1M
Cache write / 1M
1M in + 1M out$0.68
Takestext
Returnstext
Toolsyes
Reasoningoptional
Open weightsyes
Released2025-09-29
Knowledge cutoff2025-07-31
Retires
Catalog iddeepseek/deepseek-v3.2-exp

Benchmarks

sourceepoch
eci144.86
confidenceConfident
accessibilityOpen weights (unrestricted)

Benchmark scores by Epoch AI, CC BY 4.0.

Using it

In 00, this model is picked per agent — and per tier, so the model that answers your customers need not be the one that writes your code. Get 00.

← all models