modelbenchmark.io

Gemini-3-Pro

Google · gemini-3-pro

Compare

Legacy model retained for compatibility with older integrations

Specification

gemini listing

Context
1M
Max output
66K
Released
2025-10-22
Knowledge cutoff
2025-11
Retires
2026-03-09
Open weights
no
Input
text, image, video, audio
Output
text

Price

US dollars per million tokens · gemini list price

Input
$2.00
Output
$12.00
Cache read
$0.20
Cache write
$2.50

Batch and priority prices

US dollars per million tokens · per listing

PriceListing$/M
Batch inputvertex_ai-language-models$1.00
Batch outputvertex_ai-language-models$6.00
Priority inputvertex_ai-language-models$3.60
Priority outputvertex_ai-language-models$21.60

Batch input is 50% below the interactive input price at the same listing.

Long-context price tiers

the rate above each context threshold

AbovePrice$/M
200KCache read$0.40
200KCache write$0.25
200KInput$4.00
200KOutput$18.00

Rate limits

as listed

Requests per minute
2,000
Tokens per minute
800,000

Quality

6 benchmarks · 2 sources

Composite
91st
percentile of 334 scored models
Rank
31 / 334
±0.07 sd
Evidence
6 × 2
3 or more benchmarks, from 2 or more sources
Effort range
6.2
points between effort settings
coding76th1/3 bench
reasoning93rd2/4 bench
math91st3/7 bench

Every score

one row per source and configuration — nothing averaged away

BenchmarkScoreConfigurationSourceRun
FrontierMath-v158.8 ±2.8default · 2 runsEpoch AI2025-11-21
GPQA diamond92.6 ±1.7defaultEpoch AI2025-11-19
OTIS Mock AIME 2024-202591.4 ±3.7defaultEpoch AI2025-11-19
Chess Puzzles31.0 ±4.6defaultEpoch AI2025-12-08
SWE-Bench verified75.8default · mini-SWE-agent · 2 runsSWE-bench
72.9 ±2.0defaultEpoch AI2026-02-13
69.6high · mini-SWE-agentSWE-bench
Sources and settings disagree by 6.2 points. Both figures stand.
FrontierMath-Tier-4-2025-07-019.4 ±5.7default · 2 runsEpoch AI2025-11-21

The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.

Available from 16 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
geminivendor$2.00$12.00$0.201M66K2026-03-09
qihang-ai$0.57$3.431M65K
poe$1.60$9.60$0.161M66K
jiekou$1.80$10.801M66K
302ai$2.00$12.001M64K
databricks
$2.00
$2.50
$12.00
$15.00
$0.20
$0.25
$2.501M66K
gmi · gmi$2.00$12.001M66K
helicone$2.00$12.00$0.201M66K
merge-gateway$2.00$12.00$0.201M66K
opencode$2.00$12.00$0.201M66K
openrouter$2.00$12.00$0.201M66K
replicate · replicate$2.00$12.00
vertex_ai$2.00$12.00$0.201M66K
vertex_ai-language-models$2.00$12.00$0.201M66K2026-03-26
anyapi1M66K
github_copilot · github-copilot128K64K

More from Google

most-hosted first