| 厂商 | TypeSafe AI |
| 类型 | 判断模型 |
| 参数规模 | 未公开 |
| 开源/闭源 | 闭源 |
| 架构 | System One 判断架构(并行计算,非自回归) |
| 输入价格 | $0.042/M tokens |
| 输出价格 | $0/M tokens |
| 语言 | 多语言 |
| 发布时间 | 2026-09 |
| 评分 | 4.6 / 5 |
| 能力 | classification、routing、scoring、judgmentclassification、routing、scoring、judgment |
| 标签 | 判断模型、分类、路由、评分、自动化Judgment Model、Classification、Routing、Scoring、Automation |
| 优点 | 成本极低(输出永久免费)、速度极快(并行计算,毫秒级)、校准概率可量化不确定性、零类型幻觉(不编造选项)、适合自动化决策 / 路由 / 审核Extremely low cost (output free forever)、Blazing fast (parallel compute, millisecond-level)、Calibrated probabilities quantify uncertainty、Zero type-hallucination (never invents options)、Ideal for automated decisions / routing / moderation |
| 注意事项 | 不生成文本,无法对话或写作、不适合创意与开放问答、事实正确性取决于输入质量、生态与工具链较新Cannot generate text — no chat or writing、Unsuited for creative or open-ended Q&A、Factual correctness depends on input quality、Ecosystem and tooling still nascent |
Jev 目前以付费 API / 平台调用为主,可关注官方是否提供免费试用额度。Jev is mainly available via paid API/platform calls; watch for official free trial quotas.
Jev 的上下文窗口为 -,适合处理长文档与多轮复杂对话。Jev's context window is -, suitable for long documents and multi-turn complex dialogue.
闭源模型,主要通过官方 API 或平台使用。Closed-source model, mainly available via official API or platform.
Jev 的主要能力包括 classification、routing、scoring、judgment,可用于相关场景的创作、分析与自动化任务。Jev's main capabilities include classification、routing、scoring、judgment, usable for creation, analysis and automation in related scenarios.