Compare deepinfra/llama-3.2-3b-instruct vs gemini/gemini-3-pro-preview

Model Comparison

Metric	llama-3.2-3b-instruct deepinfra	gemini-3-pro-preview gemini
Pricing
Input Price	$0.01 / 1M tokens	$2 / 1M tokens
Output Price	$0.02 / 1M tokens	$12 / 1M tokens
Capabilities
Context Window	131072 tokens	1M tokens
Capabilities		tools, reasoning
Input type	text	text, image, audio, video
Category Scores
Overall Average	17.3	80.1
Maths	3.3	95.7
Finance	7.3	84.3
Science	30.5	84.7
Writing	21.2	75.1
Academia	22.0	82.1
Marketing	19.9	76.6
Programming	N/A	62.3
Benchmark Tests
HLE	5.2	37.2
AIME	6.7	_
GPQA	32.8	91.3
MMLU	63.4	_
SciCode	5.2	56.1
MATH-500	48.9	_
MMLU-Pro	34.7	89.8
LiveCodeBench	8.3	91.7
AA Math Index	3.3	95.7
AA Coding Index	_	62.3
AAII	11.2	72.8

llama-3.2-3b-instruct Advantages

• Lower input token cost ($0.01 / 1M tokens vs $2 / 1M tokens)

gemini-3-pro-preview Advantages

• Higher benchmark scores (80.1 vs 17.3)
• Larger context window (1M tokens vs 131072 tokens)

Compare deepinfra/llama-3.2-3b-instruct vs gemini/gemini-3-pro-preview