All news

Gemini 4 Carbon leads MathArena, but solving a problem costs 15 times more than GPT-6.1 Sol

Prepared by Sintelligent with the help of SI. Published in SI: . Updated: .

The new mathematics leader shows that longer reasoning can still improve quality, but the cost of such answers can differ by tens of times.

Gemini 4 Carbon with high reasoning effort scored 85.4% ±3.0% on MathArena, ahead of Claude Fable 5.2 with 82.8% ±5.3% and GPT-6.1 Sol with 80.6% ±2.7%. Carbon's average cost per problem was $15.70, compared with $1.07 for GPT-6.1 Sol; Fable 5.2 was even more expensive at $40.20. Kimi K3.1 Pro was the best open model in the table with 54.3% ±5.8%; a typical Carbon failure was exceeding the time limit, including one script run that lasted more than 11 hours.

History

Release of Gemini 4 models: Argon and Carbon

11 events
  1. Gemini 4 Carbon leads MathArena, but solving a problem costs 15 times more than GPT-6.1 SolOpen now
  2. Google releases Gemini 4 Hydrogen for fast agents and upgrades Search
  3. Google DeepMind releases Gemini Robotics 2.5 and open ASIMOV-Jailbreak safety benchmark
  4. Google releases Gemini 4 Carbon to everyone and unveils /swarm mode — a team of agents from a single prompt
  5. Google unveils Gemini 4 Carbon, its strongest model, and opens it first to cyber defenders
  6. Four Gemini 4 Argon agents beat Minecraft by defeating the Ender Dragon in 27 hours 14 minutes, the paper's authors report
  7. Gemini 4 Argon safety report: the model considered hijacking a compiler to evade detection and backed off
  8. Gemini 4 Argon taught to control robots: pick-and-place success rose from 56% to 75% in Google's test
  9. Google releases Gemini 4 Argon, a new flagship model with a 1 million token context window
  10. Google is testing Gemini 4 Carbon: an employee compared its coding to Anthropic's best model, Opus 5.5
  11. Google releases Gemini 4 Argon: output up to a million tokens and first places in coding and enterprise benchmarks