DeepSeek V4 Pro is DeepSeek's large mixture-of-experts model with 1.6 trillion total parameters and 49 billion active per token. It first appeared as a preview in April 2026, and the general availability release followed in August 2026, with the main gains in agent capability in production settings. DeepSeek offers three thinking effort levels for it: low, high and max. In Jantaris it has a 1,000,000-token context window and up to 384,000 output tokens. The weights are published as open source.
Best for
- Agentic workflows that were tuned on V4 Pro
- Long-context analysis of codebases and document sets
- Very long generated outputs such as full reports
- Comparing an open-weight model against closed ones on your own data
How to call it
Use the OpenAI SDK with the Jantaris base URL:
from openai import OpenAI
client = OpenAI(base_url="https://jantaris.com/v1", api_key="sk-jt-...")
resp = client.chat.completions.create(
model="deepseek-v4-pro",
messages=[{"role": "user", "content": "Draft a technical design doc for a job queue."}],
)
print(resp.choices[0].message.content)FAQ
How much does DeepSeek V4 Pro cost per 1M tokens?
$2.35 / $4.70 per 1M tokens in Jantaris. Output tokens cost noticeably more than input, so long generations drive most of the bill on this model.
Will DeepSeek V4 Pro be retired?
DeepSeek first announced it would redirect V4 Pro traffic to V4.1 Flash, then reversed that decision after user feedback and kept V4 Pro available. A V4.1 Pro has been mentioned but has no public date, so plan for a future migration.
Is V4 Pro better than DeepSeek V4.1 Flash?
Not by DeepSeek's own account: it says DeepSeek V4.1 Flash beats V4 Pro in tests by several parties while costing less. V4 Pro still makes sense where your existing prompts and evals were built around it.
How to call DeepSeek V4 Pro via API
from openai import OpenAI
client = OpenAI(base_url="https://jantaris.com/v1", api_key="sk-jt-...")
resp = client.chat.completions.create(
model="deepseek-v4-pro",
messages=[{"role": "user", "content": "Write a haiku about Baltic amber."}],
)
print(resp.choices[0].message.content)curl https://jantaris.com/v1/chat/completions \
-H "Authorization: Bearer $JANTARIS_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model": "deepseek-v4-pro", "messages": [{"role": "user", "content": "Write a haiku about Baltic amber."}]}'import OpenAI from "openai";
const client = new OpenAI({ baseURL: "https://jantaris.com/v1", apiKey: process.env.JANTARIS_API_KEY });
const resp = await client.chat.completions.create({
model: "deepseek-v4-pro",
messages: [{ role: "user", content: "Write a haiku about Baltic amber." }],
});
console.log(resp.choices[0].message.content);