Online

DeepSeek V4 Pro

DeepSeek · OpenAI-compatible API

deepseek/deepseek-v4-pro-0813

Create account →Quickstart ↓

USD / 1M tokens

Input$0.79

Cache Read$0.026

Cache Write—

Output$2.38

Context: 1M
Max Output: 384K
Tool Calling: ✓
Reasoning: ✓
Vision: ✗
Streaming: ✓

OpenAI-compatible API

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.iteracompute.com/v1",
    api_key=os.environ["ITERACOMPUTE_API_KEY"],
)
response = client.chat.completions.create(
    model="deepseek/deepseek-v4-pro-0813",
    messages=[{"role": "user", "content": "Hello!"}],
)
print(response.choices[0].message.content)

import OpenAI from "openai";

const client = new OpenAI({
  baseURL: "https://api.iteracompute.com/v1",
  apiKey: process.env.ITERACOMPUTE_API_KEY,
});
const response = await client.chat.completions.create({
  model: "deepseek/deepseek-v4-pro-0813",
  messages: [{ role: "user", content: "Hello!" }],
});
console.log(response.choices[0].message.content);

curl https://api.iteracompute.com/v1/chat/completions \
  -H "Authorization: Bearer $ITERACOMPUTE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "deepseek/deepseek-v4-pro-0813",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

deepseek/deepseek-v4-pro-0813View Docs →

Catalog updated: 2026-10-19.

Overview

Text input, streaming and structured outputs; up to 393,216 output tokens.

Tasks and selection tradeoffs

Complex change review

Review invariants, failure paths and test gaps in a proposed refactor.

Large text-context reasoning

Compare distant source sections and require line references for conclusions.

Task suggestions based on catalog features and price, not model quality benchmarks.

Choose your tool

OpenAI SDKCall from Python or JavaScript Vercel AI SDKBuild a TypeScript application

Choose a model and tool for your task →

Pricing

Prices in USD per 1M tokens.

Input$0.79

Cache Read$0.026

Cache Write—

Output$2.38

Example task: Review a large change and its tests

Assume 60,000 aggregate input tokens and 5,000 billable output tokens across all calls. No cache hit.

60,000 × $0.79 / 1M + 5,000 × $2.38 / 1M = $0.059300

If usage actually reports 45,000 of those input tokens as cached, replace that portion at the cached-input rate ($0.026/1M): $0.024920. Cached tokens are part of total input, never added on top.

Illustrative calculation using the 2026-10-19 catalog snapshot. Include every agent call and retry, repeated history and billable reasoning output; settle against actual usage and current Console prices.

Estimate your monthly cost

Uses the 2026-10-19 catalog snapshot, independently of the live prices above. USD per 1M tokens; an estimate, not a bill.

Estimate includes standard text input, cache reads and text output only. Long-context premiums, service tiers, cache writes, image output and tool fees are excluded.

Model

Input tokens / request

Billable output tokens / request

Cached share of input (%)

Requests / month

Input includes cached tokens; output includes billable reasoning. Count every API request, including agent steps and retries.

Estimated monthly cost

—USD / month

DeepSeek V4 Pro

Uncached input: —
Cache read: —
Billable output: —

Connect this model →

Calculation and billing basis

Monthly cost = requests × [input × (1 − cached share) × input rate + input × cached share × cache-read rate + output × output rate] / 1,000,000.

Cached tokens are part of total input, never added on top. Use actual cache-read hits from usage; repeating a prompt does not guarantee a hit. Missing rates are unconfirmed, not $0. Separate cache-write, storage and other additional fees are excluded. Settle against actual usage and current Console prices.

Enable JavaScript to use the calculator. The worked example above remains available.

Usage limits

Text-only. Input and output cost more than V4.1 Flash in this snapshot; compare solved-task cost before choosing Pro. A larger context declaration does not prove retrieval accuracy.
See the current context and maximum output in the model specifications above. Input and requested output share the context budget.
The context and capability flags describe the public catalog. They do not establish a particular tool workflow or a quality score.
Start with one representative task. Check correctness, actual usage and response time before increasing traffic.
Use the integration examples to connect your application and handle tool results.

API parameters

Parameter / protocol field	Usage
`model / messages`	Use the exact model ID and chat message list
`tools / tool_calls / tool messages`	Use tools to define functions and role=tool to return their results
`stream / stream_options.include_usage`	Use SSE streaming with final usage
`max_tokens`	Set an output budget within the model and account limits

Parameters vary by model. Keep model-specific controls separate from shared OpenAI-compatible fields.

API

Base URL: https://api.iteracompute.com/v1
Endpoint: POST /v1/chat/completions
Model ID: deepseek/deepseek-v4-pro-0813
Authentication: Authorization: Bearer ITERACOMPUTE_API_KEY

View Docs

FAQ

How do I use DeepSeek V4 Pro with the OpenAI SDK?

Set base_url to https://api.iteracompute.com/v1 and model to deepseek/deepseek-v4-pro-0813. Authenticate with your IteraCompute API key.

What does DeepSeek V4 Pro cost?

See the pricing section for current input, cached-input and output rates in USD per 1M tokens. Check the refresh time before use.

What context length is available?

See the model specifications above for the current context and maximum output. Input and requested maximum output share the context budget.

Ready for your first request?

Create your key, choose a tool and get a non-empty first reply.

Create an account →Choose a tutorial