Skip to content
AI News HubLIVE
In-site rewrite2 min read

๐Ÿš€ Introducing DeepSeek-V3 | DeepSeek API Docs

Summary

DeepSeek-V3 is the biggest leap forward yet, featuring 671B MoE parameters (37B activated), 60 tokens/second (3x faster than V2), and training on 14.8T high-quality tokens. API pricing is same as V2 until Feb 8, then $0.27/M input (cache miss), $0.07/M (cache hit), $1.10/M output. Fully open-source models and papers.

๐Ÿš€ Introducing DeepSeek-V3 | DeepSeek API Docs
Report an error

The correction channel is not available yet. You can copy the article reference below for later.

Correction instructions
Read article

๐Ÿš€ Introducing DeepSeek-V3 | DeepSeek API Docs

Skip to main content

DeepSeek API Docs

English

English

ไธญๆ–‡๏ผˆไธญๅ›ฝ๏ผ‰

DeepSeek Platform

Quick Start

Your First API Call

Models & Pricing

Token & Token Usage

Rate Limit

Error Codes

Agent Integrations

API Guides

Thinking Mode

Multi-round Conversation

Chat Prefix Completion (Beta)

FIM Completion (Beta)

JSON Output

Tool Calls

Context Caching

Anthropic API

API Reference

News

DeepSeek-V4 Preview Release 2026/04/24

DeepSeek-V3.2 Release 2025/12/01

DeepSeek-V3.2-Exp Release 2025/09/29

DeepSeek V3.1 Update 2025/09/22

DeepSeek V3.1 Release 2025/08/21

DeepSeek-R1-0528 Release 2025/05/28

DeepSeek-V3-0324 Release 2025/03/25

DeepSeek-R1 Release 2025/01/20

DeepSeek APP 2025/01/15

Introducing DeepSeek-V3 2024/12/26

DeepSeek-V2.5-1210 Release 2024/12/10

DeepSeek-R1-Lite Release 2024/11/20

DeepSeek-V2.5 Release 2024/09/05

Context Caching is Available 2024/08/02

New API Features 2024/07/25

Other Resources

FAQ

Change Log

News

Introducing DeepSeek-V3 2024/12/26

On this page

๐Ÿš€ Introducing DeepSeek-V3

Biggest leap forward yetโ€‹

โšก 60 tokens/second (3x faster than V2!)

๐Ÿ’ช Enhanced capabilities

๐Ÿ›  API compatibility intact

๐ŸŒ Fully open-source models & papers

๐ŸŽ‰ Whatโ€™s new in V3โ€‹

๐Ÿง  671B MoE parameters

๐Ÿš€ 37B activated parameters

๐Ÿ“š Trained on 14.8T high-quality tokens

๐Ÿ”— Dive deeper here:

Model ๐Ÿ‘‰ https://github.com/deepseek-ai/DeepSeek-V3

Paper ๐Ÿ‘‰ https://github.com/deepseek-ai/DeepSeek-V3/blob/main/DeepSeek_V3.pdf

๐Ÿ’ฐ API Pricing Updateโ€‹

๐ŸŽ‰ Until Feb 8: same as V2!

๐Ÿคฏ From Feb 8 onwards:

Input (cache miss)Input (cache hit)Output

$0.27/M tokens$0.07/M tokens$1.10/M tokens

Still the best value in the market! ๐Ÿ”ฅโ€‹

๐ŸŒŒ Open-source spirit + Longtermism to inclusive AGI

๐ŸŒŸ DeepSeekโ€™s mission is unwavering. Weโ€™re thrilled to share our progress with the community and see the gap between open and closed models narrowing.

๐Ÿš€ This is just the beginning! Look forward to multimodal support and other cutting-edge features in the DeepSeek ecosystem.

๐Ÿ’ก Together, letโ€™s push the boundaries of innovation!

Previous

Introducing DeepSeek App

Next

๐Ÿš€ DeepSeek V2.5: The Grand Finale ๐ŸŽ‰

Biggest leap forward yet

๐ŸŽ‰ Whatโ€™s new in V3

๐Ÿ’ฐ API Pricing Update

Still the best value in the market! ๐Ÿ”ฅ

WeChat Official Account

Community

Email

Discord

Twitter

More

GitHub

Copyright ยฉ 2026 DeepSeek, Inc.

Key points and analysis

Article intelligence

EngineersAdvanced

Key points

  • DeepSeek-V3: 671B MoE, 37B activated, 60 tokens/sec.
  • Trained on 14.8T high-quality tokens.
  • API pricing: $0.27/$0.07 per M input (cache miss/hit), $1.10 per M output after Feb 8.
  • Open-source models and papers; commitment to inclusive AGI.

Highlights and analysis are generated automatically and may contain errors. Check the original source.