Pixtral Large

Unknown Developer

This cutting-edge multimodal model, built upon Mistral Large 2, boasts 124 billion parameters and demonstrates exceptional image understanding. It expertly interprets documents, charts, and natural images, without sacrificing its strong text-based performance. Equipped with a 123 billion parameter multimodal decoder and a 1 billion parameter vision encoder, the model also features a 128K context window, allowing it to process up to 30 high-resolution images simultaneously.

Model Specifications

Technical details and capabilities of Pixtral Large

Core Specifications

124.0B Parameters

Model size and complexity

128.0K / 128.0K

Input / Output tokens

November 17, 2024

Release date

Capabilities & License

Multimodal Support
Supported
Web Hydrated
No
License
Mistral Research License (MRL) for research; Mistral Commercial License for commercial use

Resources

API Reference
https://mistral.ai/
Playground
https://chat.mistral.ai/

Performance Insights

Check out how Pixtral Large handles various AI tasks through comprehensive benchmark results.

100
75
50
25
0
93.8
AI2D
93.8
(94%)
93.3
DocVQA
93.3
(93%)
88.1
ChartQA
88.1
(88%)
80.9
VQAv2
80.9
(81%)
74
MM-MT-Bench
74
(74%)
69.4
MathVista
69.4
(69%)
64
MMMU
64
(64%)
AI2D
DocVQA
ChartQA
VQAv2
MM-MT-Bench
MathVista
MMMU

Detailed Benchmarks

Dive deeper into Pixtral Large's performance across specific task categories. Expand each section to see detailed metrics and comparisons.

Non categorized

MathVista

Current model
Other models
Avg (62.3%)

DocVQA

Current model
Other models
Avg (92.1%)

VQAv2

Current model
Other models
Avg (79.2%)

AI2D

Current model
Other models
Avg (90.5%)

MM-MT-Bench

Current model
Other models
Avg (73.5%)

Providers Pricing Coming Soon

We're working on gathering comprehensive pricing data from all major providers for Pixtral Large. Compare costs across platforms to find the best pricing for your use case.

OpenAI
Anthropic
Google
Mistral AI
Cohere

Share your feedback

Hi, I'm Charlie Palars, the founder of Deepranking.ai. I'm always looking for ways to improve the site and make it more useful for you. You can write me through this form or directly through X at @palarsio.

Your feedback helps us improve our service

Stay Ahead with AI Updates

Get insights on Gemini Pro 2.5, Sonnet 3.7 and more top AI models