Refresh Models —Deals —Regulatory —Sources —

Questions this page answers

Which is better, DeepSeek-V2 or Gemma 3 1B?: Across 8 shared benchmarks, DeepSeek-V2 scores higher on 7 and Gemma 3 1B on 1. The widest gap is MMLU-Pro, where DeepSeek-V2 scores 58.5 against 14.7.

DeepSeek-V2 vs Gemma 3 1B

Across 8 shared benchmarks, DeepSeek-V2 scores higher on 7 and Gemma 3 1B on 1. The widest gap is MMLU-Pro, where DeepSeek-V2 scores 58.5 against 14.7.

DeepSeekvsGoogle8 shared benchmarks7–1 head-to-head

Benchmark	DeepSeek-V2	Gemma 3 1B
bbh	78.9	39.1
GPQA Diamond	35.3	19.2
GSM8K	79.2	62.8
humaneval	48.8	41.5
ifeval	57.7	80.2
LiveCodeBench	20.3	1.9
MBPP	66.6	0.4
MMLU-Pro	58.5	14.7

Best tracked score per model per benchmark (default configuration; source-attributed). ↓ marks lower-is-better metrics. Open either model for its full surface, provenance and pricing.