Which is better, DeepSeek-V4.1-Flash or Gemini 3.1 Pro?

DeepSeek-V4.1-Flash vs Gemini 3.1 Pro

26 SHARED BENCHMARKS

Across 26 shared benchmarks, DeepSeek-V4.1-Flash scores higher on 14 and Gemini 3.1 Pro on 12. The widest gap is terminalbenchV40, where DeepSeek-V4.1-Flash scores 26.8 against 4. DeepSeek-V4.1-Flash is the cheaper of the two on tracked API pricing ($0.30 against $2.00 per million input tokens).

DEEPSEEKVSGOOGLE26 SHARED14–12 HEAD-TO-HEAD

BenchmarkDeepSeek-V4.1-FlashMARGINGemini 3.1 Pro

AA Intelligence4046.5

AA-LCR8482

AA-Omniscience-5.332.9

baby_vision89.651.6

critpt14.317.7

cybergym88.138.8

DeepSWE 1.174.212

gdpval56.623.1

GPQA Diamond90.994.3

HLE39.251.4

HLE (with tools)63.951.6

livebench_agentic_coding77.344.1

livebench_coding8076.5

livebench_data_analysis79.378.5

livebench_instruction_following7079.1

livebench_language81.285.4

livebench_math93.391

livebench_reasoning86.784

MMMU-Pro7783

nl2repo6433.4

OmniScience Accuracy46.455.3

OmniScience Non-Hallucination3.550.1

Program Bench20.339.5

scicode51.959

Terminal-Bench 2.190.674

terminalbenchV4026.84

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. Quoted rates are the price-setter row we currently track for each model — its direct or vendor-official listing where one exists (direct, google-official), otherwise the lowest tracked offer.

Compare Gemini 3.1 Pro withALL PAIRINGS →

GoogleGemini 3 Pro Gemma 4 26B A4B Gemini 2.5 Pro Gemini 1.5 Pro Gemma 4 31B Gemma 4 E2B

Across the fieldGPT-5.5 Claude Opus 4.8 GPT-4o GPT-5.6 Sol GPT-5.4 GPT-5.2 Claude Opus 4.6 Claude Sonnet 4.5