GoogleReleased 2026-03

Gemini 3.1 Pro

Google DeepMind science and long-context leader. Highest GPQA Diamond in this snapshot with a 2M-token context window.

Context

Speed

85 t/s

Price /M

$2.5 / $15

Access

api

When to use it

Scientific reasoning, huge document corpora, and Google Cloud / Workspace stacks.

Coding Arena trails Claude/OpenAI leaders; verify agent harness fit.

ScienceLong contextMultimodalAgents

Arena Elo1492

Human preference ranking from blind pairwise chats. Higher is better; top frontier models cluster within ~50–80 Elo.

AA Intelligence Index54

Artificial Analysis composite across agents, coding, science, and general evaluations (v4.1 weighting).

SWE-bench Verified80.6%

Percent of real GitHub issues resolved end-to-end. Strong signal for agentic coding usefulness.

GPQA Diamond94.3%

PhD-level science questions. Separates frontier reasoning models better than saturated knowledge tests.

MMLU-Pro92.6%

Harder multi-choice knowledge/reasoning suite than classic MMLU.

Humanity's Last Exam44.7%

Frontier closed-ended academic difficulty across many domains.