The Decoder ↗ · 2026-11-15 00:48 · UTC+8
Gemini 4 Carbon wins MathArena championship, but costs 15 times more than GPT-6.1 Sol
Independent The Decoder Source language: English
Google's Gemini 4 Carbon clinched first place in the MathArena math benchmark, surpassing GPT-6.1 Sol, but its testing cost is about 15 times that of the latter. Anthropic's Claude Fable 5.2 placed second, and Moonshot AI's Kimi K3.1 Pro was the top open-source model.
Full text
Short News
Copy the url to clipboardShare this articleGo to comment section
Gemini 4 Carbon takes the MathArena crown but costs 15 times more than GPT-6.1 Sol
Matthias BastianView the LinkedIn Profile of Matthias Bastian
Nov 14, 2026
Google's Gemini 4 Carbon has become the first non-OpenAI model in a while to top the MathArena math benchmark. It knocked GPT-6.1 Sol from the top spot, while Claude Fable 5.2 came in second, making it potentially one of the most capable publicly available models for math tasks. OpenAI's models, which recently allegedly solved demanding equations like the Navier-Stokes equations, aren't publicly available yet.
Gemini 4 Carbon's most common failure mode is exceeding the time limit. The model also almost never gives up on a task, a trait that can be read as either positive or negative. In one case, it launched a script that ran for more than eleven hours without checking on its progress. That extra performance comes at a price, though. With the reasoning settings used in testing, Carbon and Fable 5.2 cost roughly 15 and 40 times as much as GPT-6.1 Sol, respectively.
RecommendationModelProviderRelease dateExpected performanceExpected cost
#1 overallGemini 4 Carbon (high)GoogleNovember 10, 202685.4% ±3.0%$15.70 ±$1.33
#2 overallClaude-Fable-5.2 (max)AnthropicNovember 9, 202682.8% ±5.3%$40.20 ±$2.59
#3 overallGPT-6.1 Sol (max)OpenAISeptember 29, 202680.6% ±2.7%$1.07 ±$0.018
Best open modelKimi K3.1 Pro (max)Moonshot AIOctober 23, 202654.3% ±5.8%$3.70 ±$0.27