D

DeepSeek: deepseek-v3-0324

deepseek-v3-0324
128K konteks8.2K keluaranPenalaranAlatFile
Dirilis Mar 24, 2025Batas pengetahuan Aug 2025Diperbarui Aug 20, 2026

A state-of-the-art 671B MoE model (37B active) featuring Multi-head Latent Attention (MLA) and auxiliary-loss-free load balancing. It provides a hybrid "Think/Non-think" mode, delivering SOTA performance in coding, math, and agentic tasks while remaining the most cost-efficient flagship in the 2026 market.

Mode chatDeprecation Jul 2026

Harga

Harga input
$0.50/ 1 J token
Harga output
$2.00/ 1 J token
Jendela konteks 163.8K tokenEndpoint kompatibel openaiVendor DeepSeek

Waktu aktif

Performa

Memuat data performa...

Penggunaan & Peringkat

Memuat penggunaan...

Pertanyaan yang sering diajukan

Model serupa