法律 AI 团队终于有了可落地的降本增效方案——路由+微调开源模型比纯用前沿模型便宜 11 倍还更准，做垂直领域 AI 应用的开发者可以直接参考 Harvey 的实践。
Hugging Face CEO Clement Delangue 指出，通过路由（routing）和后训练（post-training）开源模型，不仅能提升系统准确性，还能显著降低成本和提高速度。法律 AI 公司 Harvey 与 Fireworks AI 合作，在 100 个法律任务上测试发现：混合智能体（GLM 5.1 为主，路由到 Opus 4.7 作为顾问）在质量和成本上均优于单一前沿模型，18% 全通过率 vs 14%，成本 $368 vs $954。后训练（SFT）将 Kimi 2.6 的全通过率从 11% 提升至 15%，超过 Opus 的 14%，成本仅 $84，是 Opus 的 1/11。这挑战了“前沿模型在所有场景都更好”的营销叙事，为开发者提供了更务实、经济的 AI 系统构建思路。
Routing and post-training open-source models won't only give you more accurate systems but also mean...
Routing and post-training open-source models won't only give you more accurate systems but also meaningfully faster and cheaper systems as most companies are currently learning (in addition to giving you more control and privacy). The idea that a "frontier" model (by frontier we mean is slightly more accurate on a few very limited benchmarks) will be better for all domains, all tasks, all setups just doesn't hold up! It's marketing for making you pay more! Harvey @harvey We partnered with @FireworksAI_HQ to train open-source models for legal. Here's what we found: 1) Hybrid legal agents can beat frontier models on quality and cost by routing selectively to a frontier advisor. We tested a hybrid setup where GLM 5.1 served as the primary worker, routing tasks to Opus 4.7 as an advisor when needed. GLM invoked Opus sparingly, just 0.83 times per task on average. The hybrid setup beat Opus on both quality and cost: 18% all-pass vs 14%, at $368 vs $954 across the same 100 tasks. 2) Post-training can push open models to frontier-level legal performance. On a 100-task slice of our Legal Agent Benchmark (LAB), SFT moved Kimi 2.6's all-pass rate from 11% to 15%, beating Opus' 14%. But the cost gap was even more striking: $84 vs $954 across the same 100 tasks, or ~11x cheaper. We're excited to continue working with @FireworksAI_HQ on the next generation of open-source legal agents. 🔗 View Quoted Tweet 💬 3 🔄 0 ❤️ 4 👀 193 📊 3 ⚡