Muse Wonder

Meta's models gained 4 to 8 index points in six weeks, with Terminal-Bench up 24 points. Wonder Full is limited to Muse Code users, and cost per task is high.

• articleBenchmarking Gpt 6 Astra
• articleArtificial Analysis Capability Indices V1 1
• articleAa 3d
• articleClaude Haiku 5 5

• postArtificial Analysis: Muse Wonder Full scores 56, agentic evals jumpArtificialAnlys
• postMuse Spark 1.2 places Meta on the Cost per Task Pareto frontier, scoring 6…Artificial Analysis
• postMuse Spark 1.3 scores 64 on coding tasks at a fifth of Opus 5's cost per taskArtificial Analysis

Checking sign-in…

Loading comments…