B+
Artificial Analysis (opens in a new tab)
Operator：Artificial AnalysisMulti-dimensional dataLive
Methodology (opens in a new tab)Links
EVALUATION CATALOG
AI evaluations
Browse model capability, preference, agent, and real-world usage evaluations by platform and benchmark.
8
11 entries are omitted until an English presentation is available.
Platforms
6- B+
OpenRouter Rankings (opens in a new tab)
Operator：OpenRouterMulti-dimensional dataLive dataMethodology (opens in a new tab)Links10 more links
- Benchmarks (opens in a new tab)
- Models by task (opens in a new tab)
- Fastest models (opens in a new tab)
- Cost per session (opens in a new tab)
- Languages (opens in a new tab)
- Programming (opens in a new tab)
- Context length (opens in a new tab)
- Tool calls (opens in a new tab)
- Images (opens in a new tab)
- Top apps (opens in a new tab)
- B+
Arena.ai (opens in a new tab)
Operator：Arena IntelligenceMultimodal arenaLiveMethodology (opens in a new tab)Links - B+
Epoch AI Capabilities & Benchmarks (opens in a new tab)
Operator：Epoch AIResearch data platformLiveMethodology (opens in a new tab) - B-
Scale Labs / SEAL (opens in a new tab)
Operator：Scale AIExpert evaluation platformLiveMethodology (opens in a new tab)Links - A-
Stanford HELM (opens in a new tab)
Operator：Stanford CRFMMulti-scenario evaluation suiteLiveMethodology (opens in a new tab)
Benchmarks
2- C+
CursorBench (opens in a new tab)
Operator：Cursor / AnysphereCoding / Agent3.2Methodology (opens in a new tab) - B+
Terminal-Bench (opens in a new tab)
Operator：Terminal-Bench teamCoding / Agent2.xMethodology (opens in a new tab)