DE

DeepSeek-V4-Pro

Reasoning modelDeepSeek ProDeepSeek V4

DeepSeek V4 Pro

Release date: 2026-04-24Updated: 2026-06-15 07:18:21.641Knowledge cutoff: 2025-059,605
Parameters
1.6T
Context length
1M
Chinese support
Supported
Reasoning ability

DeepSeek V4 Pro is an AI model published by DeepSeek-AI, released on 2026-04-24, for Reasoning model, with 1.6T parameters, and 1M context length, under the MIT License license, with a 3206.00 score on CodeForces.

Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology

DeepSeek-V4-Pro

Model basics

Reasoning traces
Supported
Thinking modes
Thinking Level · Max (Default)Standard ModeThinking Level · High
Context length
1M tokens
Max output length
375K tokens
Model type
Reasoning model
Modality (in / out)
Text, Image → Text
Release date
2026-04-24
Model file size
No data
MoE architecture
Yes
Total params / Active params
1.6T / 49B
Knowledge cutoff
2025-05
DeepSeek-V4-Pro

Open source & experience

Code license
Weights license
MIT License- 免费商用授权
GitHub repo
GitHub link unavailable
DeepSeek-V4-Pro

Official resources

DataLearnerAI blog
No blog post yet
DeepSeek-V4-Pro

API details

API speed
4/5
💡Default unit: $/1M tokens. If vendors use other units, follow their published pricing.
Standard
TypeConditionInputOutput
Text-$0.435/ 1M$0.870/ 1M
Cache PricingPrompt Cache
TypeTTLWriteRead
Text1d$0.0036/ 1M$0.870/ 1M
DeepSeek-V4-Pro

Benchmark Results

DeepSeek-V4-Pro currently shows benchmark results led by LiveCodeBench (1 / 120, score 93.50), MMLU Pro (11 / 126, score 87.50), SWE-bench Verified (10 / 108, score 80.60). This page also consolidates core specs, context limits, and API pricing so you can evaluate the model from benchmark results and deployment constraints together.

Thinking
Tool usage

General Knowledge

12 evaluations
Benchmark / mode
Score
Rank/total
GPQA Diamond
Standard Mode
72.90
103 / 179
89.10
25 / 179
90.10
21 / 179
MMLU Pro
Standard Mode
82.90
46 / 126
87.10
13 / 126
87.50
11 / 126
LiveBench
Standard Mode
73.58
23 / 115
HLE
Standard Mode
7.70
143 / 159
HLE
High
34.50
66 / 159
HLE
HighTools
44.70
36 / 159
HLE
Max
37.70
57 / 159
HLE
Extra-HighTools
48.20
30 / 159

Coding and Software Engineer

14 evaluations
Benchmark / mode
Score
Rank/total
2919
4 / 16
3206
2 / 16
LiveCodeBench
Standard Mode
56.80
75 / 120
89.80
6 / 120
93.50
1 / 120
SWE-bench Verified
Standard ModeTools
73.60
41 / 108
79.40
18 / 108
SWE-bench Verified
Extra-HighTools
80.60
10 / 108
SWE-bench Multilingual
Standard ModeTools
69.80
15 / 20
74.10
6 / 20
SWE-bench Multilingual
Extra-HighTools
76.20
5 / 20
SWE-Bench Pro - Public
Standard ModeTools
52.10
29 / 44
54.40
22 / 44
SWE-Bench Pro - Public
Extra-HighTools
55.40
19 / 44

常识推理

1 evaluations
Benchmark / mode
Score
Rank/total
Simple Bench
Standard Mode
50.90
28 / 63

AI Agent - Information Search

2 evaluations
Benchmark / mode
Score
Rank/total
BrowseComp
HighTools
80.40
12 / 45
BrowseComp
Extra-HighTools
83.40
9 / 45

AI Agent - Tool Usage

3 evaluations
Benchmark / mode
Score
Rank/total
Terminal Bench 2.0
Standard ModeTools
59.10
22 / 46
63.30
14 / 46
Terminal Bench 2.0
Extra-HighTools
67.90
9 / 46

Math and Reasoning

3 evaluations
Benchmark / mode
Score
Rank/total
IMO-AnswerBench
Standard Mode
35.30
20 / 20
88
5 / 20
89.80
3 / 20

Productivity Knowledge

1 evaluations
Benchmark / mode
Score
Rank/total
GDPval-AA
Extra-HighTools
1554
4 / 21

Compare with other models

DeepSeek-V4-Pro

Publisher

DeepSeek V4 Pro

Model Overview

DeepSeek V4 Pro is an AI model published by DeepSeek-AI, released on 2026-04-24, for Reasoning model, with 1.6T parameters, and 1M context length, under the MIT License license, with a 3206.00 score on CodeForces.

DataLearner on WeChat

Follow DataLearner on WeChat for AI model updates and research notes.

DataLearner WeChat QR code