← Back to case files
CASE FILE / Model & agent routing
Inspired by @typesafeai , here is a Jev-compatible public API to play wi…
Inspired by @typesafeai , here is a Jev-compatible public API to play with It runs a comparable open model (Qwen3.6-35B-A3B), and just uses SGLang radix cache to preserve the prefill reuse / really fast parallel systemone generation - 64 tasks in <1s. https://t.co/bbRh4yAPAB https://t.co/misKQE8n8f
● ● ●CASE FILEGUIDE
Eric Zhang · @ekzhang1
Original post
Inspired by @typesafeai , here is a Jev-compatible public API to play with It runs a comparable open model (Qwen3.6-35B-A3B), and just uses SGLang radix cache to preserve the prefill reuse / really fast parallel systemone generation - 64 tasks in <1s. https://t.co/bbRh4yAPAB https://t.co/misKQE8n8f
Views 109,684Likes 873Reposts 50
Chinese translation
受@typesafeai的启发，这是一个可供使用的与Jev兼容的公共API 它运行类似的开放模型（Qwen 3.6 - 35 B-A3 B），并且仅使用SGLang基数缓存来保留预填充重用/非常快速的并行系统生成-64个任务在1秒内。 https://t.co/bbRh4yAPAB https://t.co/misKQE8n8f
What this case shows
- Let JEV choose the model or tool first
- Auto-route on high confidence, escalate otherwise
- This is the most common production pattern