Kimi K3 is ranked 3rd on artificial analysis, only 2 points behind Sol
3 hours ago
- #Agent Evaluation
- #Benchmarking
- #AI Models
- Independent analysis of AI models and providers to help choose the best for specific use cases.
- Personalized model recommender for recommendations based on intelligence, speed, and cost priorities.
- Exploration of AI agents for general work, coding, customer support, and other tasks with comparisons across capabilities, pricing, and platform support.
- Premium plans offering expanded benchmark data, custom visualizations, and industry reports.
- Artificial Analysis Intelligence Index evaluating leading AI models, with sub-indexes for cost per task and frontier language model intelligence over time.
- Coding Agent Index measuring performance, cost, and execution time for leading coding agents on end-to-end software engineering tasks.
- Image, video, and speech leaderboards with 95% confidence intervals from arenas like Text to Image and Text to Speech.
- Agentic Index and Intelligence Evaluations covering agentic real-world work tasks, tool use, coding, reasoning, knowledge, and long-horizon tasks.
- Specific benchmarks: AA-Briefcase for long-horizon knowledge work, AA-Omniscience for knowledge and hallucination, GDPval-AA v2 for economically valuable tasks.
- Artificial Analysis Openness Index assessing model openness based on availability and transparency components.
- Output Tokens, Price and Cost, and Speed & Latency evaluations for leading AI models.