DeepSeek-OCR-2 API.
Make documents usable.

DeepSeek-OCR-2 is listed on Blade for document OCR. Extract readable content from document images so your application can search, validate and route the information.

€5 in free credits for every new customer.

Try DeepSeek-OCR-2

From €0.0232 per million input tokens. View the pricing details.

Document OCR
with DeepSeek-OCR-2.

Practical starting points for DeepSeek-OCR-2. Choose one task, define the expected output and measure the result on your own examples.

Digitize scanned pages

Prepare readable page images and evaluate extracted text against the source. Preserve page identifiers for later review.

Prepare documents for retrieval

Turn validated OCR output into chunks for an embedding model. Keep the original page alongside each indexed passage.

Assist data-entry workflows

Extract candidate values for invoices or forms, then validate dates, totals and required fields in your own application.

Explore the document ocr workflow ↗

DeepSeek-OCR-2 pricing

DeepSeek-OCR-2
API pricing.

These are catalogue “From” rates in EUR. Dynamic pricing can change the applicable price. Confirm your model, request route and active billing unit before estimating a workload.

See DeepSeek-OCR-2 on the rate card ↗

Input / 1M tokens: From€0.0232
Output / 1M tokens: From€0.0232

A concrete example
10,000 input tokens + 2,000 output tokens = €0.000278 at the displayed starting rates.

Illustration only; excludes retries, other models and paid tools. Token rates do not establish a per-page parsing price. Catalogue updated 2026-10-22.

Check current rates ↗

API integration

How to use
the DeepSeek-OCR-2 API.

Use the Blade model ID deepseek-ocr-2 and keep your API key in your backend environment. Verify account access and supported parameters before connecting production traffic.

Create a Blade API key in the console.
Run pip install openai and set BLADE_API_KEY.
Check the model list below, then use the console’s confirmed inference example for this model.

Read the API reference ↗

python

import os
from openai import OpenAI

client = OpenAI(
    base_url="https://api.models.blade.sh/v1",
    api_key=os.environ["BLADE_API_KEY"],
    timeout=120.0,
)

model_id = "deepseek-ocr-2"
models = client.models.list()
match = next(
    (item for item in models.data
     if item.id == model_id),
    None,
)
if match is None:
    raise RuntimeError("Model not listed for this account")
print(match.model_dump_json(indent=2))

Install openai and set BLADE_API_KEY. This checks the account’s model list; it does not run inference. Copy the confirmed model-specific inference request from the console before integrating.

DeepSeek-OCR-2 API FAQ.

What can I use DeepSeek-OCR-2 for?

DeepSeek-OCR-2 is listed on Blade for document OCR. Extract readable content from document images so your application can search, validate and route the information. Its Blade model ID is deepseek-ocr-2.

How much does DeepSeek-OCR-2 cost on Blade?

Input / 1M tokens: from €0.0232. Output / 1M tokens: from €0.0232. These are dynamic catalogue starting rates as of 2026-10-22, not fixed prices for every request. Check the console for current rates.

What should I confirm before integrating?

The published parsing reference does not explicitly map DeepSeek-OCR-2 to its supported parsing heads. Its catalogue token rates do not establish a per-page price. Confirm the route, output schema and billing unit in the console.

Model reference: Publisher model reference ↗. Hosted configuration and pricing: Blade catalogue, 2026-10-22.

Choose by task, input type and billing unit. Browse all models ↗

DeepSeek-OCR-2 API.Make documents usable.

Document OCRwith DeepSeek-OCR-2.