Skip to main content
AIsa gives you OpenAI-compatible access to the current DeepSeek model family with one API key, unified billing, usage logs, and gateway routing. deepseek-v4-pro exposes POST /v1/messages, POST /v1/chat/completions, and POST /v1/responses. Other DeepSeek routes in the current catalog use POST /v1/chat/completions.

​
Supported DeepSeek models

Model IDContextCapabilitiesInput / 1MOutput / 1MBest for
deepseek-r1262,144Coding, Text$0.4018$1.6058Reasoning-heavy tasks
deepseek-v3262,144Coding, Text$0.2009$0.8029Low-cost general and coding work
deepseek-v3.1262,144Coding, Text$0.4018$1.2047Balanced reasoning and coding
deepseek-v3.2128,000Coding, Text$0.2009$0.3017Very low-cost general use
deepseek-v4-flash262,144Coding, Text$0.0980$0.1960Lowest-cost high-throughput DeepSeek route
deepseek-v4-pro262,144Coding, Text$0.3045$0.6090Stronger DeepSeek route with multiple endpoint styles

​
Quickstart

from openai import OpenAI

client = OpenAI(api_key="YOUR_AISA_API_KEY", base_url="https://api.aisa.one/v1")

response = client.chat.completions.create(
    model="deepseek-v4-flash",
    messages=[{"role": "user", "content": "Review this pull request for security and correctness."}]
)
print(response.choices[0].message.content)

​
Common choices

NeedUse
Cheapest DeepSeek text/codingdeepseek-v4-flash
General current DeepSeek routedeepseek-v3.2
Reasoning-focused routedeepseek-r1
Messages or Responses endpoint compatibilitydeepseek-v4-pro

​
Data privacy

AIsa routes DeepSeek requests through an enterprise-backed aggregation path. Customer data is not used for model training under AIsa’s enterprise agreements. For compliance details, contact us.

​
Next