Google · proprietary

Gemini 3.1 Pro

Gemini 3.1 Pro by Google appears in 1 source with Reasoning at 59.1. Best read for LLM, Long context, Low cost.

CreatorGoogle

Release date2026-02-19

Knowledge cutoffNot published

Context1M tokens

Input price$2.5/M tokens

Output price$15/M tokens

Modalitytext + vision

CountryUS

Metrics

All source-backed metrics

Open in leaderboard

LLM Stats Rank2ranking · source_rank

Reasoning59.1reasoning · index_reasoning

Math55.55math · index_math

Coding43.21coding · index_code

Research37.38research · index_search

Vision40.41multimodal · index_vision

Tool calling32.38tool_calling · index_tool_calling

Long context11.98long_context · index_long_context

Finance4.5domain · index_finance

Legal4.46domain · index_legal

Healthcare29.15domain · index_healthcare

GPQA94.3 %reasoning · gpqa_score

SWE-bench Verified80.6 %coding · swe_bench_verified_score

Code Arena2,086.04 %coding · coding_arena_score

Humanity Last Exam51.4 %reasoning · hle_score

ARC-AGI v277.1 %reasoning · arc_agi_v2_score

MMMLU92.6 %reasoning · mmmlu_score

MMMU-Pro80.5 %multimodal · mmmu_pro_score

BrowseComp85.9 %research · browsecomp_score

MCP Atlas69.2 %tool_calling · mcp_atlas_score

MRCR v226.3 %long_context · mrcr_v2_score

SciCode59 %coding · scicode_score

Apex Agents33.5 %agent · apex_agents_score

SWE-bench Pro54.2 %coding · swe_bench_pro_score

Context1,048,576 tokenscontext · context

Speed128.52 c/sperformance · throughput

Latency18,962.51 msperformance · latency

Input price2.5 $/Mpricing · input_price

Output price15 $/Mpricing · output_price

Evidence

Citations and source overlap

LLM StatsPublic full LLM leaderboard advanced explorer snapshot, including ranking, benchmark indexes, pricing, speed, context, license, and model metadata.Retrieved 2026-06-01

FAQ

How should I read this profile?

Treat this as a source-backed model dossier, not an EvalKit-run verification. The public values are replicated from linked sources and kept source-scoped.

Is Gemini 3.1 Pro verified by EvalKit?

No. EvalKit currently shows 0 verified rows until real run evidence exists.

Why can metrics disagree?

Different sources test different tasks, dates, prompts, and aggregation methods. EvalKit keeps those differences visible instead of merging them into a fake universal score.