gpt-5.5-pro-pre-release
52.4%±2.9%
49.5%CI55.3%
Română · beta. Traducerea a fost făcută automat și este încă în curs de revizuire. Citește pagina în engleză
aiqbench · Benchmark-uri LLM
52.4%±2.9%
49.5%CI55.3%
51.7%±2.9%
48.8%CI54.6%
50.0%±2.9%
47.1%CI52.9%
O singură grilă, toate coloanele. Alege axe din panoul Axe — o axă clasifică precum un singur clasament, mai multe se combină într-un scor compus ponderat, toate la un loc se citesc ca indexul. Locul este atribuit înaintea oricărui filtru și nu se renumerotează niciodată.
| Model | Organizație | ± CI | Oscilație | Adaugă la comparație | |||||
|---|---|---|---|---|---|---|---|---|---|
| 1 | gpt-5.5-pro-pre-release | OpenAI | — | — | 52.4% | 49.5%–55.3% | #23–60 | ||
| 2 | gpt-5.5-pre-release | OpenAI | — | — | 51.7% | 48.8%–54.6% | #22–36 | ||
| 3 | gpt-5.4-pro-2026-03-05 | OpenAI | mar. 2026 | $30 / $180 | 50.0% | 47.1%–52.9% | #20–171 | ||
| 4 | gpt-5.4-2026-03-05 | OpenAI | sept. 2026 | $2.50 / $15 | 47.6% | 44.7%–50.5% | #12–29 | ||
| 5 | claude-opus-4-8 | Anthropic | sept. 2026 | $5 / $25 | 47.2% | 44.3%–50.2% | #15–33 | ||
| 6 | claude-opus-4-7 | Anthropic | sept. 2026 | $5 / $25 | 43.8% | 40.9%–46.7% | #1–18 | ||
| 7 | claude-opus-4-6 | Anthropic | sept. 2026 | $5 / $25 | 40.7% | 37.8%–43.6% | #2–25 | ||
| 8 | gpt-5.2-2025-12-11 | OpenAI | dec. 2025 | $1.75 / $14 | 40.7% | 37.8%–43.6% | #34–92 | ||
| 9 | muse-spark | Meta AI | sept. 2026 | — | 39.0% | 36.1%–41.9% | #16–40 | ||
| 10 | gemini-3.5-flash | Google DeepMind | sept. 2026 | $1.50 / $9 | 39.0% | 36.1%–41.8% | #10–26 | ||
| 11 | kimi-k2.6 | Moonshot | apr. 2026 | $0.65 / $3.41 | 39.0% | 36.1%–41.8% | #29–78 | ||
| 12 | gemini-3-pro-preview | Google DeepMind | — | — | 37.6% | 34.8%–40.4% | #19–120 | ||
| 13 | gemini-3.1-pro-preview | Google DeepMind | sept. 2026 | $2 / $12 | 36.9% | 34.1%–39.7% | #9–36 | ||
| 14 | gemini-3-flash-preview | Google DeepMind | — | $0.50 / $3 | 35.6% | 32.8%–38.5% | #27–99 | ||
| 15 | glm-5.1 | Z.ai (Zhipu AI) | apr. 2026 | $0.97 / $3.04 | 33.4% | 30.7%–36.2% | #44–75 | ||
| 16 | gpt-5-2025-08-07 | OpenAI | aug. 2025 | $1.25 / $10 | 32.4% | 29.7%–35.2% | #64–149 | ||
| 17 | claude-sonnet-4-6 | Anthropic | sept. 2026 | $3 / $15 | 32.4% | 29.6%–35.2% | #38–77 | ||
| 18 | gpt-5.1-2025-11-13 | OpenAI | nov. 2025 | $1.25 / $10 | 31.0% | 28.3%–33.8% | #45–242 | ||
| 19 | gemini-2.5-deep-think-2025-08-01-webapp | Google,Google DeepMind | — | — | 29.0% | 26.3%–31.7% | — | ||
| 20 | gpt-5.4-mini-2026-03-17 | OpenAI | oct. 2026 | $0.75 / $4.50 | 28.3% | 26.2%–30.3% | #75–180 | ||
| 21 | fireworks/kimi-k2p5 | Moonshot | — | — | 27.9% | 25.3%–30.5% | #80–115 | ||
| 22 | gpt-5-mini-2025-08-07 | OpenAI | aug. 2025 | $0.25 / $2 | 27.2% | 24.6%–29.9% | #126–343 | ||
| 23 | qwen3.6-plus | Alibaba | oct. 2026 | $0.33 / $1.95 | 26.2% | 23.6%–28.8% | #47–265 | ||
| 24 | gpt-5.4-nano-2026-03-17 | OpenAI | oct. 2026 | $0.20 / $1.25 | 25.9% | 23.3%–28.4% | #95–222 | ||
| 25 | o4-mini-2025-04-16 | OpenAI | apr. 2025 | $1.10 / $4.40 | 24.8% | 22.3%–27.4% | #129–274 | ||
| 26 | qwen3.6-max-preview | Alibaba | oct. 2026 | — | 23.1% | 20.6%–25.6% | #34–53 | ||
| 27 | fireworks/deepseek-v3p2 | DeepSeek | — | — | 22.1% | 18.0%–26.2% | — | ||
| 28 | moonshotai/Kimi-K2-Thinking | Moonshot | — | — | 21.4% | 19.0%–23.8% | — | ||
| 29 | qwen3.5-plus | Alibaba | feb. 2026 | $0.30 / $1.80 | 21.0% | 18.6%–23.4% | #106–403 | ||
| 30 | claude-opus-4-5-20251101 | Anthropic | nov. 2025 | $5 / $25 | 20.7% | 18.3%–23.1% | #20–62 | ||
| 31 | grok-4-0709 | xAI | iul. 2025 | — | 19.7% | 17.3%–22.0% | #95–276 | ||
| 32 | o3-2025-04-16 | OpenAI | apr. 2025 | $2 / $8 | 18.7% | 16.4%–21.0% | #79–318 | ||
| 33 | glm-5 | Z.ai (Zhipu AI) | feb. 2026 | $1 / $3.20 | 16.4% | 14.2%–18.6% | #57–153 | ||
| 34 | claude-sonnet-4-5-20250929 | Anthropic | sept. 2025 | $3 / $15 | 15.2% | 13.1%–17.3% | #66–232 | ||
| 35 | gemini-2.5-pro | Google DeepMind | iun. 2025 | $1.25 / $10 | 14.1% | 12.1%–16.2% | #60–432 |
Se afișează FrontierMath exact așa cum a publicat-o sursa — propriul scor și interval de încredere, fără mix. Selectează o a doua axă ca să construiești un scor compus ponderat.