Back to Models
deepseek/deepseek-v4-pro
DeepSeek V4 Pro
DeepSeek V4 Pro is a large-scale Mixture-of-Experts model from DeepSeek with 1.6T total parameters and 49B activated parameters, supporting a 1M-token context window. It is designed for advanced reasoning, coding,...
4/24/2026
1,048,576 tokens
#15 Text (Japanese)
Input: ¥77.741/M
Output: ¥155.482/M
Specifications
Modalities
Input
text
Output
text
Supported Parameters
frequency_penalty
include_reasoning
logit_bias
logprobs
max_tokens
min_p
presence_penalty
reasoning
reasoning_effort
repetition_penalty
response_format
seed
stop
structured_outputs
temperature
tool_choice
tools
top_k
top_logprobs
top_p
Max Output Tokens
384,000Code Examples
curl https://api.fastmetal.ai/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "deepseek-v4-pro",
"messages": [{"role": "user", "content": "Hello!"}]
}'Leaderboard
Text
🏆OverallELO: 1,457
#46🇯🇵JapaneseELO: 1,467
#15🇨🇳ChineseELO: 1,501
#40🇰🇷KoreanELO: 1,422
#35🇬🇧EnglishELO: 1,465
#46frenchELO: 1,481
#35germanELO: 1,470
#23spanishELO: 1,453
#43russianELO: 1,458
#43💻CodingELO: 1,501
#51🧮MathELO: 1,445
#56✍️Creative WritingELO: 1,442
#39📝Instruction FollowingELO: 1,453
#39🌶️Hard PromptsELO: 1,480
#43💬Multi-TurnELO: 1,472
#39Code
🏆OverallELO: 1,446
#35