Google logo

Gemini 4 Neon

Multimodal
Google

Gemini 4 Neon is Google DeepMind's fast Gemini 4 reasoning model for complex software engineering, knowledge work, and science workflows, generally available on the Gemini API paid tier on 2026-11-18 (model id gemini-4-neon). Google positions it between Gemini 4 Hydrogen and Gemini 4 Carbon. It is not available on the free tier. It uses a 1,048,576-token unified context window with up to 64K default synchronous output tokens and long decode up to 1M tokens, accepts text, image, video, audio, and PDF input with text output, and supports four thinking levels: low, medium, high (default), and xhigh. Gemini API paid pricing is flat across the 1M window at $3.00 per million input tokens, $15.00 per million output tokens (including thinking tokens), and $0.15 per million implicitly cached input tokens; explicit caching is not supported. Vertex AI global endpoints use the same $3.00 / $15.00 / $0.15 rates; non-global endpoints are billed at a 10% premium. The knowledge cutoff is March 2026.

Key Specifications

Parameters
-
Context
1.0M
Release Date
November 18, 2026
Average Score
75.2%

Timeline

Key dates in the model's history
Announcement
November 18, 2026
Last Update
November 20, 2026
Today
November 22, 2026

Technical Specifications

Parameters
-
Training Tokens
-
Knowledge Cutoff
-
Family
-
Capabilities
MultimodalZeroEval

Pricing & Availability

Input (per 1M tokens)
$3.00
Output (per 1M tokens)
$15.00
Max Input Tokens
1.0M
Max Output Tokens
1.0M
Supported Features
Function CallingStructured OutputCode ExecutionWeb SearchBatch InferenceFine-tuning

Benchmark Results

Model performance metrics across various tests and benchmarks

Other Tests

Specialized benchmarks
Agents' Last Exam
Agents' Last Exam v1.1. Gemini 4 Neon model card evaluation table (results as of November 2026); pass@1, xhigh thinking unless noted. • Self-reported
75.8%
AutomationBench
Zapier AutomationBench. Gemini 4 Neon model card evaluation table (results as of November 2026); pass@1, xhigh thinking unless noted. • Self-reported
56.3%
Chartography (no tools)
Chartography (tool setting not specified in model card). Gemini 4 Neon model card evaluation table (results as of November 2026); pass@1, xhigh thinking unless noted. • Self-reported
73.5%
FrontierCode 2.0
FrontierCode 2.0 main score at default high thinking setting. Gemini 4 Neon model card evaluation table (results as of November 2026); pass@1, xhigh thinking unless noted. • Self-reported
46.2%
Graphwalks BFS >128k
GraphWalks 256k to 1M subset, BFS (F1). Gemini 4 Neon model card evaluation table (results as of November 2026); pass@1, xhigh thinking unless noted. • Self-reported
95.2%
LVBench
LVBench. Gemini 4 Neon model card evaluation table (results as of November 2026); pass@1, xhigh thinking unless noted. • Self-reported
89.0%
RiemannBench
RiemannBench. Gemini 4 Neon model card evaluation table (results as of November 2026); pass@1, xhigh thinking unless noted. • Self-reported
84.0%
SEC-bench Pro
SEC-bench Pro. Gemini 4 Neon model card evaluation table (results as of November 2026); pass@1, xhigh thinking unless noted. • Self-reported
90.5%
Terminal-Bench-Science 0.1
Terminal-Bench-Science 0.1. Gemini 4 Neon model card evaluation table (results as of November 2026); pass@1, xhigh thinking unless noted. • Self-reported
70.5%
Vals Index
Vals Index v2.2. Gemini 4 Neon model card evaluation table (results as of November 2026); pass@1, xhigh thinking unless noted. • Self-reported
70.9%

License & Metadata

License
proprietary
Announcement Date
November 18, 2026
Last Updated
November 20, 2026

Compare Gemini 4 Neon

All comparisons

Similar Models

All Models

Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.