Models · Google DeepMind

Gemini 3.1 Pro

flagship Closed 2026-02-19

Lab: Google DeepMind
Released: 2026-02-19
Parameters: unknown
Context window: 1M
Input: text, image, audio, video
Output: text
Weights: Closed
Official page: deepmind.google ↗

Benchmark scores

ARC-AGI 2 reasoning 77.1%

77.1% 2026-02-19 Google releases Gemini 3.1 Pro with improved reasoning

BrowseComp agentic 85.9%

85.9% 2026-04-23 OpenAI releases GPT-5.5 with agentic capabilities and doubled API pricing

HealthBench general 42.6%

42.6% 2026-08-13 VITA clinical RAG matches or outperforms frontier LLMs on HealthBench
42.6% 2026-08-13 Corpus-specific VITA RAG matches or beats frontier LLMs on HealthBench
42.6% 2026-08-13 VITA clinical RAG system matches or outperforms frontier LLMs on HealthBench

Related articles