Hormouz AIHormouzSign in

DeepSeek: DeepSeek V3.1: AI model

DeepSeek V3.1 is a large hybrid-reasoning model (671B parameters, 37B active) with switchable thinking and non-thinking modes. Tuned for coding, tool use and agentic work, via Hormouz.

Context window:
163,840 tokens
Input:
$0.25/1M tokens
Output:
$0.95/1M tokens
Type:
text
  • Tools
  • Streaming
  • Long context

Thinking levels:

DeepSeek V3.1 is a large hybrid-reasoning model with 671 billion parameters (37 billion active per pass) that supports two response modes: a "thinking mode" for step-by-step reasoning and a "non-thinking mode" for faster replies. The reasoning behaviour can be controlled with the reasoning parameter.

It succeeds DeepSeek V3 0324 and improves tool use, code generation and reasoning efficiency. Reaching performance close to DeepSeek R1 on some hard benchmarks while responding more quickly. Its context window runs up to about 160K tokens and it supports tool calling and structured output, making it a fit for research, coding and agentic workflows.

Pricing is usage-based: $0.65/M in · $1.70/M out. Hormouz provides stable access to the model from inside Iran, with no foreign credit card or circumvention tooling required.

DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context...

Quickstart

curl https://api.hormouz.net/v1/chat/completions \
  -H "Authorization: Bearer ng-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-chat-v3.1",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Specs

VendorDeepSeek
TypeText
Context window163,840 tokens
Pricing$0.25/M in · $0.95/M out
APIChat Completions
CapabilitiesStreaming · Tools

Benchmarks

Reasoning

GPQA Diamond
Graduate-level scientific reasoning
77.9%
HLE
Humanity's Last Exam
14.3%
IFBench
Instruction-following
41.5%
AA-LCR
Long-context reasoning
56.7%
CritPt
Research-level physics reasoning
2%
AA Omniscience
Broad factual knowledge accuracy
29%
Non-Hallucination
Rate of avoiding fabricated answers
17.5%

Coding

Terminal-Bench Hard
Agentic coding & terminal use
25%

Agentic

τ²-Bench
Tool-use agents in dual-control scenarios
37.4%

Source: Artificial Analysis

Design Arena

How Design Arena works ↗
3D
Top 72%
1,087Elo
Win rate: 47.9%Avg time: 28.9s
Code
Top 72%
1,118Elo
Win rate: 47.8%Avg time: 45.7s
Data Viz
Top 77%
1,104Elo
Win rate: 46.2%Avg time: 34.8s
Game Dev
Top 72%
1,097Elo
Win rate: 47.1%Avg time: 48.8s
SVG
Top 99%
966Elo
Win rate: 38.2%Avg time: 7.8s
UI Component
Top 74%
1,092Elo
Win rate: 47.1%Avg time: 28.9s
Website
Top 70%
1,129Elo
Win rate: 48%Avg time: 47.9s

Metrics sourced from Design Arena

Code example

cURL
curl https://api.hormouz.net/v1/chat/completions \
  -H "Authorization: Bearer ng-YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-chat-v3.1",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'
Python
from openai import OpenAI
client = OpenAI(api_key="ng-YOUR_API_KEY", base_url="https://api.hormouz.net/v1")
response = client.chat.completions.create(
    model="deepseek/deepseek-chat-v3.1",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)
JavaScript
import OpenAI from "openai";
const client = new OpenAI({ apiKey: "ng-YOUR_API_KEY", baseURL: "https://api.hormouz.net/v1" });
const response = await client.chat.completions.create({
  model: "deepseek/deepseek-chat-v3.1",
  messages: [{ role: "user", content: "Hello!" }],
});
console.log(response.choices[0].message.content);

Buy the DeepSeek: DeepSeek V3.1 API, paid in rial

  • Buy credit with an Iranian bank card; no foreign card or account needed.
  • One Hormouz key reaches DeepSeek: DeepSeek V3.1 and 600+ other models, and you pay only for what you use.
  • No VPN and no risk of a banned account; just change the base URL to api.hormouz.net/v1.
Create an accountBuy DeepSeek API access, paid in rial

Use via Hormouz

Connect to this model and 600+ others with one API key, then see usage down to each request.

Request company access

FAQ

How much does this model cost?
Billing is usage-based: $0.65 per million input tokens and $1.70 per million output tokens. There is no fixed monthly fee. You pay only for what you use.
What are the thinking and non-thinking modes in DeepSeek V3.1?
The model has two modes: a thinking mode for step-by-step reasoning and complex tasks, and a non-thinking mode for faster replies. You control the reasoning behaviour with the reasoning parameter.
How is DeepSeek V3.1 different from DeepSeek V3 0324?
V3.1 succeeds V3 0324, improving tool use, code generation and reasoning efficiency, and adds a hybrid reasoning mode.
How do I access this model from inside Iran?
Create an account on Hormouz and generate an API key. Hormouz provides a stable path to the model from inside Iran, with no foreign credit card required.
Is this model compatible with OpenAI code?
Yes. Hormouz exposes an OpenAI-compatible API. Just change the base URL to Hormouz and swap in your Hormouz key, and your existing SDKs run without a rewrite.

Updated 2026-10-13