DeepSeek V4 Pro 0813

DeepSeek V4 Pro 0813 production weights with 1M context and 384K max output, MIT open-sourced.

Input price: $0.636–$1.452
Output price: $1.908–$4.356

Context: 1M
Max output: 384K

Capability: 156 (#36 of 79)
Human preference: 1445
30-day uptime: —

InputTextOutputText

Providers

Providers that serve this model. With Auto, each request goes to the best available provider. Prices in USD per 1M tokens; stats cover the last 30 days.

AutoInput$0.636–$1.452Output$1.908–$4.356Cache read$0.0219–$0.143Context1MMax output384KWe send each request to the best-priced healthy provider. You pay that provider’s price.Alibaba Cloud (Model Studio Intl)CheapestInput$0.636–$1.272Output$1.908–$3.816Cache read$0.064–$0.127Global (HK/DE/JP)No trainingDeepSeekInput$0.657–$1.314Output$1.971–$3.942Cache read$0.0219–$0.0438ChinaNo trainingTogether AIInput$1.452Output$4.356Cache read$0.143Context1MUnited Statesfp4No trainingNo retention

Pricing

Prices vary by provider. Showing Alibaba Cloud (Model Studio Intl), the lowest. Pick a provider above to see its prices.

Input: $0.636; per 1M tokens
Output: $1.908; per 1M tokens
Cache read: $0.064; per 1M tokens

01:00–04:00, 06:00–10:00 UTC

Input: $1.272; per 1M tokens
Output: $3.816; per 1M tokens
Cache read: $0.127; per 1M tokens

Try it

Runs on your balance and is billed like any API request.

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

import os

from openai import OpenAI

client = OpenAI(
    api_key=os.environ["OHMYGPT_API_KEY"],
    base_url="https://api.ohmygpt.com/v1",
)

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({
  apiKey: process.env.OHMYGPT_API_KEY,
  baseURL: "https://api.ohmygpt.com/v1",
});

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);

The answer streams in here.

Billed by usage

Similar models

Other models with the same output types at a similar price.

Better value: more capable, lower price
Pareto frontier
This model
Similar models

Capability scores are from Epoch AI’s Epoch Capabilities Index (CC BY 4.0, as of 2026-11-10), which combines dozens of benchmarks with GPT-5 set at 150. epoch.ai/eci

Price: our current input price per 1M tokens in USD, lowest across channels. Dated snapshots and resold copies of the same model are merged into one point; free models are not shown.

Shaded area: capability above and price below the median of all models. Pareto frontier: no other model is both cheaper and more capable.

DeepSeek V4 Pro 0813This model

Capability: 156
Input: $0.636–$1.452
Output: $1.908–$4.356

DeepSeek V3.1

Capability: 141
Input: $0.638
Output: $1.848

GLM-5

Capability: 146
Input: $0.6424–$1.10
Output: $2.8908–$3.52

Kimi K2 Thinking

Capability: 146
Input: $0.66
Output: $2.75

Claude Sonnet 5 (Reseller)

Capability: 157
Input: $0.60
Output: $3.00

Human preference ratings are from the LMArena Text Arena (CC BY 4.0, as of 2026-11-09): Elo ratings from blind votes, on a different scale from capability scores. lmarena.ai/leaderboard/text