# Gemini 3.1 Pro Preview vs GPT-5.4 Mini

As of Oct 15, 2026, Gemini 3.1 Pro Preview leads GPT-5.4 Mini in 7 of 7 shared categories: General LLM, Reasoning, Science, Coding › Web dev, Long context, Factuality, and Abstraction (ARC).

Gemini 3.1 Pro Preview: $4.5 per 1M blended tokens.

GPT-5.4 Mini: $1.6875 per 1M blended tokens.

Gemini 3.1 Pro Preview: closed weights.

GPT-5.4 Mini: closed weights.

- Which to pick, General LLM: As of Oct 15, 2026, Gemini 3.1 Pro Preview leads GPT-5.4 Mini in General LLM at #15 against #59.
- Which to pick, Reasoning: As of Oct 15, 2026, Gemini 3.1 Pro Preview leads GPT-5.4 Mini in Reasoning at #14 against #59.
- Which to pick, Science: As of Oct 15, 2026, Gemini 3.1 Pro Preview leads GPT-5.4 Mini in Science at #7 against #52.
- Which to pick, Coding › Web dev: As of Oct 15, 2026, Gemini 3.1 Pro Preview leads GPT-5.4 Mini in Coding › Web dev at #58 against #74.
- Which to pick, Long context: As of Oct 15, 2026, Gemini 3.1 Pro Preview leads GPT-5.4 Mini in Long context at #12 against #71.
- Which to pick, Factuality: As of Oct 15, 2026, Gemini 3.1 Pro Preview leads GPT-5.4 Mini in Factuality at #18 against #54.
- Which to pick, Abstraction (ARC): As of Oct 15, 2026, Gemini 3.1 Pro Preview leads GPT-5.4 Mini in Abstraction (ARC) at #2 against #36.

As of Oct 15, 2026, Gemini 3.1 Pro Preview is state of the art on best per dollar in Math (since Oct 14, 2026; sources Humanity's Last Exam; methodology v1.0).

As of Oct 15, 2026, GPT-5.4 Mini is ranked #36 in Abstraction (ARC) (since Oct 15, 2026; sources ARC-AGI; methodology v1.0).