HyperAI
Command Palette
Search for a command to run...
LLM Leaderboard
Rank language models by Intelligence Index against cost, speed and latency
Data from Artificial Analysis
Scatter plot of Intelligence Index against cost per task in US dollars on a logarithmic scale. Models on the cost frontier, from cheapest: GPT-6 Luna (Low) ($0.0051, 9.7); GPT-6 Luna (Medium) ($0.042, 18.5); Claude Haiku 5.5 (Low) ($0.044, 18.6); GPT-6.1 Terra (Low) ($0.071, 29.4); Muse Ember (Xhigh) ($0.12, 34.1); Muse Ember (Max) ($0.14, 36.0); GPT-6.1 Terra (Medium) ($0.17, 38.2); GPT-6.1 Terra (High) ($0.26, 42.4); GPT-6.1 Sol (Medium) ($0.36, 46.3); GPT-6.1 Sol (High) ($0.59, 50.3); GPT-6.1 Sol (Xhigh) ($0.76, 52.5); GPT-6.1 Sol (Max) ($1.27, 54.6); GPT-6 Astra (High) ($3.24, 54.8); GPT-6 Astra (Xhigh) ($3.66, 55.3); Claude Fable 5.2 (Medium, Default Fallback) ($3.80, 56.4); Claude Opus 5.5 (Xhigh, Default Fallback) ($6.05, 57.4); Gemini 4 Carbon (Medium) ($6.59, 58.3); Gemini 4 Carbon (High) ($7.76, 60.0); Claude Fable 5.2 (High, Default Fallback) ($9.08, 61.8); Claude Fable 5.2 (Xhigh, Default Fallback) ($14.83, 62.2); Claude Fable 5.2 (Max, Default Fallback) ($26.54, 64.6)
104 of 104 shown