Gemini 2.5 Pro vs Gemini 3 Pro

Benchmarks, API pricing and specs, head to head. Data updated 2026-06-10.

Gemini 2.5 Pro

Google · Mar 2025

70.9

Google's 2025 workhorse flagship — first mainstream thinking model with a 1M context, still widely deployed.

Gemini 3 Pro

Google · Nov 2025

73.6

Google's November 2025 frontier breakout — 91.9% GPQA Diamond and 37.5% HLE made it the reasoning leader of its generation.

The verdict

Gemini 3 Pro wins 5 of the 5 benchmarks these models share, against 0 for Gemini 2.5 Pro. Gemini 2.5 Pro is about 1.3x cheaper per blended million tokens (3:1 input:output mix). Gemini 2.5 Pro also takes 1.0M of context versus 1M for Gemini 3 Pro.

Benchmark head-to-head 05

SWE-bench Verified
63.8%76.2%
Gemini 2.5 ProGemini 3 Pro
GPQA Diamond
86.4%91.9%
Gemini 2.5 ProGemini 3 Pro
AIME 2025
88%95%
Gemini 2.5 ProGemini 3 Pro
HLE
21.6%37.5%
Gemini 2.5 ProGemini 3 Pro
MMMU
~79.6%~87%
Gemini 2.5 ProGemini 3 Pro

Specs & pricing

Gemini 2.5 ProGemini 3 Pro
modhub Index70.973.6
Input price / 1M$1.25$2
Output price / 1M$10$12
Context window1.0M1M
Max output66K66K
Open weightsnono
Reasoning modelyesyes
Multimodal inputtext, image, audio, videotext, image, audio, video
Knowledge cutoffJan 2025Jan 2025
ReleasedMar 2025Nov 2025
Example monthly cost*$27.50$38.00

* 10M input + 1.5M output tokens per month at list prices, no caching. Green = better value on that row.

Frequently asked questions

Which is better, Gemini 2.5 Pro or Gemini 3 Pro?
Gemini 3 Pro wins 5 of the 5 benchmarks these models share, against 0 for Gemini 2.5 Pro. Gemini 2.5 Pro is about 1.3x cheaper per blended million tokens (3:1 input:output mix). Gemini 2.5 Pro also takes 1.0M of context versus 1M for Gemini 3 Pro.
Which is cheaper, Gemini 2.5 Pro or Gemini 3 Pro?
Gemini 2.5 Pro costs $1.25/$10 per million input/output tokens, while Gemini 3 Pro costs $2/$12. For a typical workload of 10M input and 1.5M output tokens per month, that's $27.50 versus $38.00.
Which model is better for coding, Gemini 2.5 Pro or Gemini 3 Pro?
On SWE-bench Verified — the standard agentic-coding benchmark — Gemini 3 Pro scores 76.2% versus 63.8% for Gemini 2.5 Pro, making Gemini 3 Pro the stronger pick for coding agents.

More comparisons