Google

Gemini 4 Neon

Models.dev
Intelligence
61.4
Rank 6 of 813 measured models
Output speed
105 t/s
Rank 174 of 476 measured models
Web Search
$14 /1K calls
Context window
1M
Rank 120 of 1420 measured models
Cost per task
$6.44
Rank 236 of 255 measured models

Gemini 4 Neon (Xhigh) is a text AI model from Google, released on Nov 18, 2026, with a context window of 1M tokens. Gemini 4 Neon (Xhigh) scores 61.4 on the Artificial Analysis Intelligence Index, ranking 6th of 813 measured models. The closest models on this index are GPT-6.2 Astra (62.1), Gemini 4 Carbon (60.0), and Claude Opus 5.5 (Default Fallback) (59.1). With an output speed of 105 t/s, it is faster than GPT-6.2 Astra (48.6 t/s) and Gemini 4 Carbon (58.3 t/s), slower than Claude Opus 5.5 (Default Fallback) (113 t/s). Every benchmark side by side: Gemini 4 Neon (Xhigh) vs GPT-6.2 Astra.

Artificial Analysis Indices

Extra high
Intelligence61.4
Rank 6 of 813 measured models

Reasoning levels

Each point is a reasoning level of this model: more reasoning often raises the score but costs more. Select a point to open that level.
Intelligence Indexmost efficient ↗Cost per task

Capability indexes

Performance in specific industries, measured by Artificial Analysis
Finance & Accounting65.1
Rank 1 of 283 measured models
Strategy & Ops63.8
Rank 3 of 283 measured models
Legal63.8
Rank 1 of 283 measured models
Healthcare & Medical60.1
Rank 2 of 210 measured models
Engineering66.8
Rank 4 of 281 measured models
Economics61.9
Rank 4 of 289 measured models

Standard Benchmarks

FrontierCode 2.044.4%
Rank 4 of 37 measured models
Terminal-Bench-Science70.0%
Rank 3 of 37 measured models
HLE-Diamond71.3%
Rank 3 of 37 measured models
CritPt v2 (Hard Subset)60.5%
Rank 10 of 33 measured models
AutomationBench-AA79.3%
Rank 1 of 37 measured models
GDP.pdf (all-pass)28.0%
Rank 10 of 37 measured models
FrontierMath Tiers 1–3 (Epoch AI)91.2%
Rank 5 of 82 measured models
FrontierMath Tier 4 (Epoch AI)97.6%
Rank 7 of 69 measured models
Additional benchmarks5
Design Arena · 3d76.2%
Design Arena · Codecategories65.5%
Design Arena · Gamedev64.6%
Design Arena · Uicomponent58.1%
Design Arena · Website59.9%

Models at a similar level

The closest models on the Intelligence Index, one configuration per family.

Preferences and analytics

Rybbit: visits, performance, clicks, copies, forms and session replay (masked inputs), with your permission. Refusing does not limit the site.

Privacy

Allow navigation analytics?

With your permission, Rybbit measures visits, performance and interactions (clicks, copies, forms), including session replay, to improve BenchSift. Input values are masked in replays. Refusing does not limit the site.

Your choice is stored for 180 days. Change it at any time using “Privacy preferences” in the footer. Withdrawal reloads the page.

No choice stored: Rybbit is disabled.

Privacy