DeepSeek V4 Flash
ChatDeepSeek-V4-Flash
DeepSeek V4 Flash on FlexAI: DeepSeek LLM, MIT license, served via the OpenAI-compatible Token Factory at the live market-tracked rate.
Recommended for
Pricing
Input
$0.09 / M tokens
Output
$0.179 / M tokens
Cached input
$0.0134 / M tokens
Context
768K tokens
API endpoint
/v1/chat/completions
Compatibility
OpenAI
Parameters
284B MoE (13B active)
License
MIT
Hardware
4× H100
Quantization
FP8
Estimate your monthly cost
M tokens
M tokens
Cache hit rate25%
Share of input tokens served from prompt cache: repeated system prompts, long documents, and multi-turn context.
7.5M × $0.0896/M input$0.67
2.5M × $0.0134/M cached input$0.03
2M × $0.1792/M output$0.36
Estimated monthly cost$1.06
Saved vs. no caching$0.2 (16%)
Estimate only, at the current market-tracked rate. Usage-based; no minimums.
Get an API keyQuick Start
DeepSeek-V4-Flash
from openai import OpenAI
client = OpenAI(
base_url="https://tokens.flex.ai/v1",
api_key="your-api-key",
)
response = client.chat.completions.create(
model="DeepSeek-V4-Flash",
messages=[
{"role": "user", "content": "Hello!"}
],
)
print(response.choices[0].message.content)