Aime
Measured Jul 10, 2026Source
Score
0.05
Models14 0 
Llama 3.2 Instruct 90B (Vision)
MetaMetaUnited States
This is the largest multimodal instruction-tuned model in the Meta Llama 3.2 series, featuring 90 billion parameters and support for both image and text inputs. It excels at visual understanding and complex reasoning tasks, making it suitable for sophisticated applications requiring the processing of both images and text.
MultimodalReasoning
Input / 1M tokens
$1.38Output / 1M tokens
$1.38Output tokens/s
59.81First-token seconds
0.63sSupported plans
0Benchmark history
Evaluations
Math 500
Measured Jul 10, 2026Source
Score
0.63
Scicode
Measured Jul 10, 2026Source
Score
0.24
Livecodebench
Measured Jul 10, 2026Source
Score
0.21
Hle
Measured Jul 10, 2026Source
Score
0.05
Gpqa
Measured Jul 10, 2026Source
Score
0.43
Mmlu Pro
Measured Jul 10, 2026Source
Score
0.67
Artificial Analysis Intelligence Index
Measured Jul 10, 2026Source
Score
6.2
TAU2
Measured Jul 10, 2026Source
Score
0.21
Terminalbench Hard
Measured Jul 10, 2026Source
Score
0.01
Lcr
Measured Jul 10, 2026Source
Score
0.12
Ifbench
Measured Jul 10, 2026Source
Score
0.3
Aime 25
Measured Jul 10, 2026Source
Score
0.03
Artificial Analysis Math Index
Measured Jul 10, 2026Source
Score
3.3
Plan availability
Products and plans that support this model
No products or plans have been linked to this model yet.
User ratings
Loading ratings...
Discussion
Thinking... Make sure you are connected to GitHub server