Advertise here
#43

Gemini 4 Carbon

Google Release: 2026-11-12 Tested on: 2026-11-13 01:08 google/gemini-4-carbon::medium
~1.2T total (~20B active)MoEClosedEstimated
(high) (medium) (low)

Gemini 3.8 Flash still tops Gemini 4 Argon, Carbon and Hydrogen in our tests →

Summary

Gemini 4 Carbon scores 9.2 on AI BENCHY and ranks #43. It has 8.6 reliability, a 95.7% pass rate, $1.933 total cost, and 38.78s average response time.

What makes Gemini 4 Carbon unique: It stands out most in Domain specific, where it ranks #2, while Combined is its weakest area at #16.

Model facts

Researched on 2026-11-12

Estimated
Parameters
~1.2T total (~20B active)
Architecture
MoE
Availability
Closed
License
-

Best estimate from public evidence; the vendor did not disclose every value. Google does not disclose size or architecture for this frontier Gemini 4 checkpoint. The 1.2T total / 20B active MoE values are a low-confidence estimate carried from the Gemini 4 Argon catalog entry, not vendor-reported. API-only. OpenRouter declares mandatory reasoning with low, medium and high efforts, medium default. The fallback release date records this onboarding request date.

Parameter source Availability source

Score

9.2

Consistency

8.9

Reliability

8.6

Total Cost (Current Price)

$1.933

Total Output Tokens

131,948

Total Input Tokens

306,396

Input Price

$2.000 / 1M

Output Price

$10.000 / 1M

Cache Read Price

$0.100 / 1M

Cache Write Price

$0.188 / 1M

Cache prices apply to input tokens. Reads reuse cached prompts; writes store them and can cost extra. Output tokens use the output price.

Tests Correct

Wrong Tests: 3

Attempt pass rate: 95.7%

Flaky tests

3

Flaky tests had mixed outcomes across runs (at least one pass and one fail).

Response Time (avg)

38.78s

Response Time (max): 206.61s

Response Time (total): 892.02s

API error: 2 Wrong answer: 1

Hamster playing table tennis

Prompt: Create a detailed SVG illustration of a hamster playing table tennis.

#43 Gemini 4 Carbon

medium
Cost
$0.203
Time
336.6s
Tokens
20,353 tok

Charts

Choose the first model, then click a second model to open a side-by-side page.

Total Output Tokens

Score vs Total Output Tokens

Quick Compare

Gemini 4 CarbonmediumvsMuse WondermediumGemini 4 CarbonmediumvsGPT-5.5mediumGemini 4 CarbonmediumvsClaude Opus 5.5mediumGemini 4 CarbonmediumvsClaude Sonnet 5.5maxGemini 4 CarbonmediumvsMuse Spark OpenhighGemini 4 CarbonmediumvsQwen3.8 MAXmediumGemini 4 CarbonmediumvsGemini 4 HydrogenlowGemini 4 CarbonmediumvsDeepSeek V4.1 Flash 1017lowGemini 4 CarbonmediumvsMuse EmberhighGemini 4 CarbonmediumvsGemini 4 Hydrogenxhigh

Category Breakdown

Category Score Consistency Tests Correct
Agentic 10.0 10.0
Anti-AI Tricks 10.0 10.0
Coding 10.0 10.0
Combined 4.7 1.6
Data parsing and extraction 10.0 10.0
Domain specific 8.2 7.2
General Intelligence 10.0 10.0
Instructions following 10.0 10.0
Puzzle Solving 10.0 10.0
Tool Calling 10.0 10.0
Trivia 10.0 10.0

Compared models