Alibaba leads trader consensus for third place on the Text Arena math leaderboard by year-end, with Z.ai, Meta, and Anthropic clustered closely behind amid a tight contest among mid-tier labs. Current arena data places Google and Anthropic in the top two spots on math prompts, leaving Alibaba’s Qwen3.8-Max (and the newly released Qwen4-Max) positioned just ahead of Z.ai’s GLM-5.3-Flash and Meta’s Muse Spark series. Recent releases, including Alibaba’s large MoE model with strong reasoning benchmarks and ongoing arena voting momentum for efficient Chinese models, support Alibaba’s edge, while Z.ai and Meta benefit from competitive pricing and open-weight accessibility that accelerates community testing. With roughly six weeks until resolution, further model drops or shifts in human preference votes could easily reorder the group, reflecting the fluid nature of arena-based math evaluations.
Eksperymentalne podsumowanie AI odwołujące się do danych Polymarket. To nie jest porada handlowa i nie ma wpływu na rozstrzyganie tego rynku. · Zaktualizowano1H6H1D7D30DALL
View resolved






















Uważaj na linki zewnętrzne.
Uważaj na linki zewnętrzne.
Często zadawane pytania