Research
Gemini 4 Carbon takes the MathArena crown but costs 15 times more than GPT-6.1 Sol
The Decoder · By Matthias Bastian ·
The Brief
- Google's Gemini 4 Carbon now tops the MathArena AI math benchmark.
- Anthropic's Fable 5.2 also ranks above GPT-6.1 Sol, so no OpenAI model leads for the first time in a while.
- Both models cost significantly more than GPT-6.1 Sol; the Decoder reports Gemini 4 Carbon costs 15 times more. In testing, Gemini ignored the time limit and almost never gave up.
Why it matters
If you pick models for math-heavy work, the top benchmark score now comes with a much higher price, so cost per solved problem matters more than the leaderboard alone.
Who’s affected
- Teams choosing AI models for math-heavy workloads
- Gemini 4 Carbon
- GPT-6.1 Sol
- Fable 5.2
- #ai
- #benchmarks
- #llm
- #math
- #gemini
- #pricing
For the first time in a while, no OpenAI model leads the MathArena AI math benchmark. Google's Gemini 4 Carbon and Anthropic's Fable 5.2 have knocked GPT-6.1 Sol from the top, though both cost significantly more. Gemini ignored the time limit in testing and almost never gave up…