V3.1-Terminus Comparison Testing
As an experimental version, although DeepSeek-V3.2-Exp has been validated for effectiveness on public evaluation sets, it still requires broader and larger-scale testing in real user scenarios to identify potential issues in certain long-tail use cases. To facilitate comparative testing by users, we have temporarily retained additional API access interfaces for V3.1-Terminus.
Users can simply modify the base_url
to "https://api.deepseek.com/v3.1_terminus_expires_on_20251015"
to access V3.1-Terminus, with pricing consistent with V3.2-Exp. This endpoint will remain available until October 15, 2025, 15:59 UTC.
We sincerely encourage users to provide valuable feedback during comparative testing via the following link:
https://feedback.deepseek.com/dsa
How to Conduct Comparison Testing
You can control which model version to access by modifying the base_url
:
- When using the original method to access the API, you will reach the DeepSeek-V3.2-Exp model
- When you set
base_url="https://api.deepseek.com/v3.1_terminus_expires_on_20251015"
, you are accessing theDeepSeek-V3.1-Terminus
model.
The correspondence between base_url
settings and specific model versions is shown in the table below:
API Type | base_url Setting | Model Version |
---|---|---|
OpenAI | https://api.deepseek.com | DeepSeek-V3.2-Exp |
Anthropic | https://api.deepseek.com/anthropic | DeepSeek-V3.2-Exp |
OpenAI | https://api.deepseek.com/v3.1_terminus_expires_on_20251015 | DeepSeek-V3.1-Terminus |
Anthropic | https://api.deepseek.com/v3.1_terminus_expires_on_20251015/anthropic | DeepSeek-V3.1-Terminus |
Usage Examples
Accessing V3.1-Terminus via OpenAI-Compatible API
- curl
- python
- nodejs
Invoke The API
curl https://api.deepseek.com/v3.1_terminus_expires_on_20251015/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer ${DEEPSEEK_API_KEY}" \
-d '{
"model": "deepseek-chat",
"messages": [
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "Hello!"}
],
"stream": false
}'
Sample Output
{
... ...
"model": "deepseek-v3.1-terminus",
"choices": [
{
"index": 0,
"message": {
"role": "assistant",
"content": "Hello! How can I help you today?"
},
"logprobs": null,
"finish_reason": "stop"
}
],
... ...
}
Invoke The API
# Please install OpenAI SDK first: `pip3 install openai`
import os
from openai import OpenAI
client = OpenAI(
api_key=os.environ.get('DEEPSEEK_API_KEY'),
base_url="https://api.deepseek.com/v3.1_terminus_expires_on_20251015")
response = client.chat.completions.create(
model="deepseek-chat",
messages=[
{"role": "system", "content": "You are a helpful assistant"},
{"role": "user", "content": "Hello"},
],
stream=False
)
print(f"Model is: {response.model}")
print(f"Output is: {response.choices[0].message.content}")
Sample Output
Model is: deepseek-v3.1-terminus
Output is: Hello! How can I help you today?
Invoke The API
// Please install OpenAI SDK first: `npm install openai`
import OpenAI from "openai";
const openai = new OpenAI({
baseURL: 'https://api.deepseek.com/v3.1_terminus_expires_on_20251015',
apiKey: process.env.DEEPSEEK_API_KEY,
});
async function main() {
const completion = await openai.chat.completions.create({
messages: [
{ role: "system", content: "You are a helpful assistant." },
{ role: "user", content: "Hello!."}
],
model: "deepseek-chat",
});
console.log("Model is:", completion.model)
console.log("Output is:", completion.choices[0].message.content);
}
main();
Sample Output
Model is: deepseek-v3.1-terminus
Output is: Hello! How can I help you today?
As shown in the sample output, you can verify whether the called model is V3.1-Terminus by checking the model
field in the API response.
Accessing V3.1-Terminus via Claude Code
When setting up Claude Code environment variables, you need to modify the ANTHROPIC_BASE_URL
environment variable to access the DeepSeek-V3.1-Terminus model:
export ANTHROPIC_BASE_URL=https://api.deepseek.com/v3.1_terminus_expires_on_20251015/anthropic
For complete configuration instructions, please refer to the Anthropic API Guide.