Solar Pro 4 has achieved an 89.1% score on the GPQA benchmark, demonstrating strong performance in graduate-level question answering. The model also exhibited an efficiency of 59.1 tokens per second with 62.3 intelligence points per dollar, indicating a favorable balance of speed and cost-effectiveness. These results were independently measured and are available for comparison against other models. AI
IMPACT Demonstrates improved performance and efficiency in graduate-level question answering, potentially influencing future model development.
RANK_REASON The item reports on benchmark results for an AI model, which falls under research. [lever_c_demoted from research: ic=1 ai=1.0]
Read on Mastodon — mastodon.social →
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →