Home/OCR/Mistral OCR 3
VERIFIED BY CODESOTADec 2025

Mistral OCR 3: Independent Benchmark Results

We ran the full OmniDocBench (1,355 images) and OCRBench v2 (7,400 English samples) benchmarks ourselves. Here's what we found.

Quick Install

pip install mistralai

API access via api.mistral.ai | Python SDK available | Pay-per-use pricing

Documentation

Independently Verified Benchmarks

CodeSOTA ran two full benchmark suites: OmniDocBench (Dec 19, 2025) and OCRBench v2 (Dec 21, 2025). We processed 8,755 images total through Mistral's OCR API and computed metrics using official evaluation tools.

OmniDocBench: 1,355 imagesOCRBench v2: 7,400 samplesOfficial eval toolsReproducible
90.1%
Text Accuracy
Verified
70.9%
Table TEDS
Verified
78.2%
Formula Accuracy
Verified
91.6%
Reading Order
Verified

OmniDocBench Results (Verified)

OmniDocBench is a comprehensive document parsing benchmark with 1,355 pages across 9 document types. The composite score formula: ((1-TextEditDist)*100 + TableTEDS + FormulaCDM) / 3

MetricMistral OCR 3GPT-4oPaddleOCR-VL
Composite Score79.75verified~8592.86
Text Edit Distance0.099verified0.020.03
Table TEDS70.9%verified-93.5%
Table Structure TEDS75.3%verified--
Formula (Edit Distance)0.218verified--
Reading Order91.6%verified--

Lower is better for Edit Distance metrics. CodeSOTA verification date: December 19, 2025.

OCRBench v2 Results (Verified)

Important: Mistral OCR is a pure OCR model - it extracts text/markdown from images. OCRBench v2 tests 21 task types including VQA, chart parsing, and structured extraction, which require a Vision-Language Model (VLM). Mistral OCR excels at text extraction tasks but cannot answer questions about images.

25.2%
Overall Score
7,400 samples
79.1%
Full-Page OCR
Core strength
55.2%
Doc Parsing
Verified
32.5%
Text Recognition
Verified
Task TypeScoreSamplesNotes
Full-Page OCR79.1%200Core OCR strength
APP Agent69.5%300UI text extraction
Cognition VQA58.3%800Some VQA works via text
Math QA56.2%300
Document Parsing55.2%400Structured output
Science QA44.0%300
Text Recognition32.5%800
Table Parsing4.4%400Weak on OCRBench format
Chart/Spotting/Grounding0%variousRequires VLM, not pure OCR

CodeSOTA verification date: December 21, 2025. Full results at OCRBench v2 Leaderboard.

Context: Top VLMs score 55-62% on OCRBench v2 (Seed1.6-Vision: 62.2%, GPT-4o: 47.6%). Mistral OCR's 25.2% reflects that it's an OCR tool, not a VLM - comparing these is like comparing a scanner to a research assistant.

Performance by Document Type

Mistral OCR 3 performs best on academic papers and exam papers, struggles with newspapers:

Document TypeText AccuracyTable TEDS
Academic Literature97.9%83.0%
Exam Papers92.8%88.0%
Books93.9%82.7%
Research Reports95.8%82.0%
Magazines97.9%71.0%
PPT Slides95.7%72.6%
Newspapers67.0%58.3%

Performance by Language

English
94.6%

Text accuracy

Chinese
86.1%

Text accuracy

Mixed
86.2%

Text accuracy

Pricing

Standard API
$2/1000 pages

Real-time processing via API

Batch API
$1/1000 pages

50% discount for async processing

Our full benchmark run cost $2.71 for 1,355 images using the standard API.

Code Example

from mistralai import Mistral
client = Mistral(api_key="your-api-key")
# Load document
with open("document.pdf", "rb") as f:
    doc_data = base64.b64encode(f.read()).decode()
# OCR with Mistral OCR 3
response = client.ocr.process(
    model="mistral-ocr-2512",
    document={"type": "pdf", "data": doc_data}
)
# Output is markdown with HTML tables
print(response.content)

When to Use Mistral OCR 3

Excellent For
  • Academic papers (97.9% accuracy)
  • Exam papers (92.8% + 88% tables)
  • Research reports and books
  • Cost-sensitive high-volume OCR
  • English text extraction
Weak Points
  • Newspapers (67% - avoid)
  • Complex multi-column layouts
  • Chinese text (86% vs 94% English)
  • Table recognition vs PaddleOCR

Verdict

OmniDocBench
79.75
Document parsing
OCRBench v2
25.2%
Pure OCR tasks: 79.1% full-page

Mistral OCR 3 is a dedicated OCR tool, not a VLM. It excels at text extraction (79.1% full-page OCR, 90.1% text accuracy) but cannot answer questions about images or parse charts.

For document parsing, it's behind PaddleOCR-VL (92.86) but faster and cheaper. For VLM tasks, use Qwen3-Omni or GPT-4o instead.

Best use case: High-volume text extraction from PDFs/documents where you need markdown output, not image understanding.

Model ID: mistral-ocr-2512
API: docs.mistral.ai
Verified: OmniDocBench Dec 19, OCRBench v2 Dec 21, 2025 by CodeSOTA

Related Reading