Gemini 4 Neon

  • New
  • Closed weights
  • Reasoning
  • Input: audio · file · image · text · video
  • Released 2026-11
RankedEnough results for an index.
IndexAll categories, out of 100.81.9±4.4Median 50.7
CoverageIndex weight that has results.90%
SpeedOutput tokens per second.110/sMedian 76/s
First tokenSeconds to the first output token.56.5 sMedian 1.1 s
Input / 1MUS dollars per 1M input tokens.$3.00batch $1.50Median $0.60
Output / 1MUS dollars per 1M output tokens.$15.00batch $7.50 Batch price per 1M output tokens.Median $2.50
ContextMaximum tokens in one request.1.05MMedian 262K
EloLMArena rating and rank.1524±8 #3Median 1425

Details

From OpenRouter
API IDThe model ID on OpenRouter.
google/gemini-4-neon
Max outputMaximum output tokens per request.
944K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.15
Weights
—
RetiresDate OpenRouter removes the model.
—

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.8 counted81.22 / 221
CodingCode writing and repair.4 counted83.52 / 214
ReasoningLogic problems and puzzles.4 counted87.71 / 225
MultimodalTasks with images and text.1 counted72.7—
KnowledgeFacts and expert knowledge.3 counted78.34 / 266
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.None counted——
MathMath problems.3 counted75.96 / 171

Results

23 counted
BenchmarkResultThe score from the source.PlacePlace among results on this benchmark.
OTIS Mock AIME 2024-2025Math100.0%1 / 203
Chess PuzzlesReasoning83.0% ±3.81 / 93
GPQA diamond · Epoch AIKnowledge96.3% ±1.35 / 220
Bug Hunt BenchCoding63.0%1 / 30
Agent Arena steerabilityAgentic21.22 / 59
Agent Arena task outcomeAgentic19.22 / 59
Mystery Game PuzzlesReasoning85.0% ±3.63 / 83
Furniture AssemblyReasoning89.2% ±4.02 / 40
FrontierMath-Tier-4-v2-PrivateMath97.6% ±2.44 / 75
SimpleQA VerifiedKnowledge75.2% ±1.46 / 94
cursorBench41Coding65.4%1 / 12
FrontierCode 2.0Coding46.2%3 / 32
CWE-bench v2Agentic70.7%2 / 20
FrontierSWE v2Coding62.4%3 / 25
Terminal-Bench 4.1.0Agentic61.2% ±1.58 / 52
Artificial Analysis GDP.pdfAgentic28.0%8 / 23

Sources

BenchLMCC BY-NC 4.0Epoch AICC BYTerminal-BenchNo licence statedLMArenaCC BY 4.0models.devMITCheaper InferenceNo licence stated

History

13 releases
Output / 1M$15.00 → $15.00
18 Nov 202621 Nov 2026

Questions

How much does Gemini 4 Neon cost?
$3.00 per 1M input tokens, $15.00 per 1M output tokens.
How good is Gemini 4 Neon?
Its index is 81.9 out of 100. That is place 2 of 282 ranked models.
What is Gemini 4 Neon best at?
Reasoning: 87.7 out of 100, place 1 of 225.
How fast is Gemini 4 Neon?
110 output tokens per second, 56.5 s to the first token.
What is the context window of Gemini 4 Neon?
1.05M tokens.
Is Gemini 4 Neon open source?
No. Its weights are not published.
When was Gemini 4 Neon released?
November 2026.

Badge

Gemini 4 Neon on Inference360: place 2 of 282, index 81.9

Compare with

Gemini 4 Carbon80.5Claude Fable 5.284.0GPT-6.2 Astra79.5Claude Opus 5.577.1Gemini 4 Argon76.8

Lower price

Gemini 4 Carbon80.5 · $10.00/1M

More from Google

All models
Gemini 4 Hydrogen76.2Gemini Decision 1UnrankedGemini Omni 1.2 FlashUnrankedGemini Omni 1.1 FlashUnrankedGemini 4 Carbon80.5