modelbenchmark.io

Gemini 4 Argon

Google · gemini-4-argon

Compare

Google's flagship reasoning model for complex software engineering, deep math and science, and long-horizon autonomous agents

Specification

google listing

Context
1M
Max output
33K
Released
2026-10-13
Knowledge cutoff
2026-03
Retires
—
Open weights
no
Input
text, image, video, audio, pdf
Output
text

Price

US dollars per million tokens · google list price

Input
$2.00
Output
$10.00
Cache read
$0.10
Cache write
$0.375

Batch and priority prices

US dollars per million tokens · per listing

PriceListing$/M
Batch inputgemini$1.00
Batch outputgemini$5.00
Priority inputgemini$3.60
Priority inputdatabricks$4.50
Priority outputgemini$18.00
Priority outputdatabricks$22.50

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

Rate limits

as listed

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

Quality

9 benchmarks · 1 source

Composite
98th
percentile of 344 scored models
Rank
9 / 344
±0.06 sd
Evidence
9 × 1
one source, or fewer than 3 benchmarks
Effort range
—
one configuration only
reasoning98th4/4 bench
math94th3/7 bench
knowledge100th1/1 bench

Every score

one row per source and configuration — nothing averaged away

BenchmarkScoreConfigurationSourceRun
Mystery Game Puzzles64.0 ±4.8highEpoch AI2026-10-13
SimpleQA Verified76.0 ±1.4highEpoch AI2026-10-13
EBR-bench64.8 ±4.9highEpoch AI2026-10-13
Chess Puzzles59.0 ±4.9highEpoch AI2026-10-13
Furniture Assembly72.5 ±5.6highEpoch AI2026-10-13
FrontierMath-Tiers-1-3-v286.0 ±2.1highEpoch AI2026-10-13
FrontierMath-Tier-4-v278.0 ±6.5highEpoch AI2026-10-13
GPQA diamond96.4 ±1.2highEpoch AI2026-10-13
OTIS Mock AIME 2024-2025100.0 ±0.0highEpoch AI2026-10-13

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

Available from 33 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
googlevendor$2.00$10.00$0.10—1M33K—
aihubmix$2.00$10.00$0.10—1M66K—
aixy$2.00$10.00$0.10—1M33K—
cloudflare-ai-gateway$2.00$10.00$0.10—1M66K—
commandcode$2.00$10.00$0.10—1M66K—
crossmodel$2.00$10.00$0.10$2.001M66K—
gemini$2.00$10.00$0.10—1M1M—
github_copilot · github-copilot$2.00$10.00$0.10—200K66K—
github-copilot$2.00$10.00$0.10—1M66K—
google-vertex$2.00$10.00$0.10—1M66K—
kilo$2.00$10.00$0.10$0.18751M944K—
merge-gateway$2.00$10.00$0.10—1M66K—
nano-gpt$2.00$10.00$0.10$0.18751M66K—
ofox$2.00$10.00$0.10—1M66K—
opencode$2.00$10.00$0.10—1M66K—
opencode_go · opencode-go$2.00$10.00$0.10—1M66K—
opencode-go$2.00$10.00$0.10—1M66K—
openrouter$2.00$10.00$0.10$0.18751M944K—
perplexity-agent$2.00$10.00$0.10—1M66K—
prime$2.00$10.00——1M944K—
requesty$2.00$10.00$0.10—1M66K—
vercel$2.00$10.00$0.10—1M1M—
vertex_ai$2.00$10.00$0.10—1M1M—
cortecs$2.25$11.249$0.113—1M66K—
databricks$2.50$12.50$0.125$2.501M1M—
venice$2.50$12.50$0.125—1M66K—
302ai$4.00$10.00$0.20$0.3751M66K—
edenai$4.00$20.00$0.20$0.3751M66K—
llmgateway$4.00$20.00$0.20$0.3751M1M—
llmgateway-providers · google-ai-studio$4.00$20.00$0.20$0.3751M1M—
llmgateway-providers · google-vertex$4.00$20.00$0.20$0.3751M1M—
neon$4.00$20.00$0.20—1M66K—
kyma · kyma$5.40$27.00$0.27—1M66K—

More from Google

most-hosted first