deepseek/deepseek-v4-flash

DeepSeek V4 Flash

DeepSeek V4 Flash is an efficiency-optimized Mixture-of-Experts model from DeepSeek with 284B total parameters and 13B activated parameters, supporting a 1M-token context window. It is designed for fast inference and...

4/24/2026

1,048,576 tokens

#56 Text (Korean)

Input: ¥16.7635/M

Output: ¥33.5269/M

Try in Chat Get API Key

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

Code Examples

curl https://api.fastmetal.ai/v1/chat/completions \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "model": "deepseek-v4-flash",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Leaderboard

Text

🏆OverallELO: 1,436

#76

🇯🇵JapaneseELO: 1,395

#63

🇨🇳ChineseELO: 1,476

#77

🇰🇷KoreanELO: 1,400

#56

🇬🇧EnglishELO: 1,446

#75

frenchELO: 1,443

#89

germanELO: 1,425

#76

spanishELO: 1,437

#65

russianELO: 1,437

#74

💻CodingELO: 1,482

#75

🧮MathELO: 1,426

#89

✍️Creative WritingELO: 1,408

#70

📝Instruction FollowingELO: 1,428

#70

🌶️Hard PromptsELO: 1,459

#73

💬Multi-TurnELO: 1,453

#61