Fireworks AI launches FireRouter: the frontier isn't a model, it's a router

Fireworks AI 分析：前沿不是单个模型，而是路由器，FireRouter 发布

Fireworks AI benchmarked 18 models on DeepSWE: picking the right model per task beats any single model. GPT-6 Astra alone scores 74.1% at $6.52/task. An oracle router across all 18 hits 97.6% at $1.88. Open-weight models alone reach 90.3% at $1.45. 94 of 113 tasks need a model under $3; the three priciest models are the best pick on only 3 tasks. FireRouter aims to make that per-task choice before the work starts—the post doesn't yet detail how.

Why it matters: Fireworks presents a data-backed argument using 18 models on DeepSWE: task-aware routing beats the single strongest model on both accuracy (97.6% vs 74.1%) and cost ($1.88 vs $6.52). The open-source-only result of 90.3% also provides a path that doesn't depend on closed models...