Pixtral-12B
Multimodal
Zero-eval
#1MM IF-Eval
#2VQAv2
#2MM-MT-Bench
by Mistral AI
About
Pixtral-12B is a multimodal language model developed by Mistral AI. It achieves strong performance with an average score of 66.9% across 12 benchmarks. It excels particularly in DocVQA (90.7%), ChartQA (81.8%), VQAv2 (78.6%). The model shows particular specialization in general tasks with an average performance of 75.5%. It supports a 136K token context window for handling large documents. The model is available through 1 API provider. As a multimodal model, it can process and understand text, images, and other input formats seamlessly. It's licensed for commercial use, making it suitable for enterprise applications. Released in 2024, it represents Mistral AI's latest advancement in AI technology.
Pricing Range
Input (per 1M)$0.15 -$0.15
Output (per 1M)$0.15 -$0.15
Providers1
Timeline
AnnouncedSep 17, 2024
ReleasedSep 17, 2024
Specifications
Capabilities
Multimodal
License & Family
License
Apache 2.0
Benchmark Performance Overview
Performance metrics and category breakdown
Overall Performance
12 benchmarks
Average Score
66.9%
Best Score
90.7%
High Performers (80%+)
2Performance Metrics
Max Context Window
136.2KAvg Throughput
0.1 tok/sAvg Latency
1msTop Categories
general
75.5%
vision
73.9%
roleplay
68.7%
code
62.0%
math
53.0%
Benchmark Performance
Top benchmark scores with normalized values (0-100%)
Ranking Across Benchmarks
Position relative to other models on each benchmark
DocVQA
Rank #18 of 26
#15DeepSeek VL2 Small
92.3%
#16Nova Lite
92.4%
#17GPT-4o
92.8%
#18Pixtral-12B
90.7%
#19Llama 3.2 90B Instruct
90.1%
#20DeepSeek VL2 Tiny
88.9%
#21Llama 3.2 11B Instruct
88.4%
ChartQA
Rank #17 of 24
#14Llama 3.2 11B Instruct
83.4%
#15DeepSeek VL2 Small
84.5%
#16Qwen2.5-Omni-7B
85.3%
#17Pixtral-12B
81.8%
#18Phi-3.5-vision-instruct
81.8%
#19Phi-4-multimodal-instruct
81.4%
#20DeepSeek VL2 Tiny
81.0%
VQAv2
Rank #2 of 3
#1Pixtral Large
80.9%
#2Pixtral-12B
78.6%
#3Llama 3.2 90B Instruct
78.1%
MT-Bench
Rank #10 of 11
#7Llama 3.1 Nemotron Nano 8B V1
81.0%
#8Ministral 8B Instruct
83.0%
#9Mistral Small 3 24B Instruct
83.5%
#10Pixtral-12B
76.8%
#11Llama 3.1 Nemotron 70B Instruct
9.0%
HumanEval
Rank #51 of 62
#48Llama 3.1 8B Instruct
72.6%
#49Claude 3 Sonnet
73.0%
#50Grok-1.5
74.1%
#51Pixtral-12B
72.0%
#52Gemma 3 4B
71.3%
#53Phi-3.5-MoE-instruct
70.7%
#54GPT-3.5 Turbo
68.0%
All Benchmark Results for Pixtral-12B
Complete list of benchmark scores with detailed information
DocVQA DocVQA benchmark | vision | multimodal | 0.91 | 90.7% | Self-reported |
ChartQA ChartQA benchmark | general | multimodal | 0.82 | 81.8% | Self-reported |
VQAv2 VQAv2 benchmark | vision | multimodal | 0.79 | 78.6% | Self-reported |
MT-Bench MT-Bench benchmark | roleplay | text | 76.80 | 76.8% | Self-reported |
HumanEval HumanEval benchmark | code | text | 0.72 | 72.0% | Self-reported |
MMLU MMLU benchmark | general | text | 0.69 | 69.2% | Self-reported |
IFEval IFEval benchmark | code | text | 0.61 | 61.3% | Self-reported |
MM-MT-Bench MM-MT-Bench benchmark | roleplay | text | 60.50 | 60.5% | Self-reported |
MathVista MathVista benchmark | math | text | 0.58 | 58.0% | Self-reported |
MM IF-Eval MM IF-Eval benchmark | code | text | 0.53 | 52.7% | Self-reported |
Showing 1 to 10 of 12 benchmarks