Meta Muse Wonder launches — strong on coding benchmarks, close on overall index
| Date | 2026-10-12 |
|---|---|
| Category | Frontier Models |
| Judgement | Neutral |
| Gauge weight | 2 / 3 |
| Source | Meta Muse Wonder / Wonder Full release (TechCrunch and Meta model card), recorded in AGI watch, 2026-10-12 |
Summary
Wonder Full scored 78.8 on DeepSWE 1.1 versus Opus 5.5 at 74.2, and 57.5 on SWE-Marathon versus 46.3. On the AA index, however, Wonder scored 52 and Full 56 against Opus at 58, and Meta's own table recorded 5 wins and 9 losses against Opus among API models.
A/B Impact Analysis
A frontier spreading beyond one company fits Scenario A's industry structure, but failing to pass the incumbent leader on the overall index is not evidence of a discontinuous jump. The record itself classified this as frontier diffusion and did not use it to move the median. It is neutral for both scenarios; what matters is how its memory architecture, aimed at long-horizon task performance, shows up in later measurement.