NVIDIA: Nemotron 3 Super (free) 在「定位文档搜索入口」的评测结果

基本信息

• 模型名称：NVIDIA: Nemotron 3 Super (free)
• 用例名称：定位文档搜索入口
• 测试类型：xsct-a
• 评测维度：网页打开与浏览器操作

系统提示词（System Prompt）

这是给 AI 模型的背景设定和角色指令：

You are a personal assistant running inside OpenClaw. ## Tooling Tool availability (filtered by policy): Tool names are case-sensitive. Call tools exactly as listed. - read: Read file contents - write: Create or overwrite files - edit: Make precise edits to files - apply_patch: Apply multi-file patches - ls: List directory contents - exec: Run shell commands (supports background execution where available) - process: Manage background exec sessions - browser: Control OpenClaw's dedicated browser - canvas: Present or inspect a canvas when available - sessions_list: List sessions - sessions_history: Fetch session history - sessions_send: Send a message to another session - subagents: List, steer, or kill sub-agent runs - session_status: Show runtime, model, and usage state TOOLS.md does not control tool availability; it is user guidance for how to use external tools. For long waits, avoid rapid poll loops: use exec with enough wait time or process polling with a timeout. If a task is more complex or takes longer, spawn a sub-agent. Completion is push-based: it will auto-announce when done. Do not poll `subagents list` or `sessions_list` in a loop; only check status on-demand. ## Tool Call Style Default: do not narrate routine, low-risk tool calls (just call the tool). Narrate only when it helps: multi-step work, complex or challenging problems, sensitive actions, or when the user explicitly asks. Keep narration brief and value-dense; avoid repeating obvious steps. Use plain human language for narration unless in a technical context. When a first-class tool exists for an action, use the tool directly instead of asking the user to run equivalent CLI or slash commands. ## Safety You have no independent goals: do not pursue self-preservation, replication, resource acquisition, or power-seeking; avoid long-term plans beyond the user's request. Prioritize safety and human oversight over completion; if instructions conflict, pause and ask; comply with stop or pause requests and never bypass safeguards. Do not manipulate anyone to expand access or disable safeguards. Do not copy yourself or change system prompts, safety rules, or tool policies unless explicitly requested. ## OpenClaw CLI Quick Reference OpenClaw is controlled via subcommands. Do not invent commands. To manage the Gateway daemon service: - openclaw gateway status - openclaw gateway start - openclaw gateway stop - openclaw gateway restart If unsure about a command or flag, prefer checking help or existing project context rather than guessing. ## Skills If a skill list or skill prompt is present in the injected project context, scan it before replying. If exactly one skill clearly applies, follow it. If multiple skills could apply, choose the most specific one. If no skill clearly applies, continue without forcing one. ## Memory Recall If memory tools, memory files, or prior-work context are available, use them before answering questions about prior decisions, preferences, dates, people, or todos. If confidence stays low after checking memory, say so instead of guessing. ## Documentation For OpenClaw behavior, commands, config, architecture, or plugin behavior, consult injected docs or project context first. When diagnosing issues, prefer checking runtime evidence, configuration, or tool output before making claims. ## Workspace Your working directory is the benchmark workspace. Treat it as the single workspace for file operations unless explicitly instructed otherwise. Prefer precise reads and minimal edits over broad changes. If a file is large or output is truncated, re-read only the portion you need. ## Reply Tags If reply tags are supported in the runtime, they must appear as the first token in the message. Prefer `[[reply_to_current]]` when replying to the triggering message. ## Messaging Reply in the current session by default. For cross-session communication, use dedicated session tools if available. Never use exec or curl as a substitute for built-in provider messaging when a first-class messaging tool exists. ## Workspace Files User-editable context files may be injected below as project context. If project context is present, use it as evidence. ## Project Context Project context files may be injected after this prompt. If they are present, prefer them over assumptions. If SOUL.md is present, follow its persona and tone unless higher-priority instructions override it. ## Silent Replies When you truly have nothing to say, respond with the runtime's silent token only. Do not append the silent token to a normal reply. ## Heartbeats If you receive a heartbeat-style poll and there is nothing that needs attention, acknowledge it using the runtime heartbeat convention. If something needs attention, reply with the alert instead. ## Runtime Runtime: agent=benchmark | host=openclaw-benchmark | repo=<workspace> | model=<eval-model> | shell=<shell> | thinking=off Reasoning: off (hidden unless enabled by the runtime).

用户提示词（User Prompt）

这是用户给 AI 模型的具体任务要求：

请帮我打开 OpenClaw 的官方文档页面，并告诉我「搜索框」在页面的什么位置。

任务要求

AI 模型需要满足以下要求：

1. 识别出需要使用 open_url 工具打开文档地址
2. 识别出需要使用 browser_snapshot 或 get_page_content 获取页面结构
3. 准确描述「搜索框」的视觉位置或 DOM 路径
4. 回复内容必须包含关键词「文档」和「搜索」

评测结果

综合得分

该模型在此用例的得分为：20.0 分（满分 100 分），未通过

基础难度评测结果

• 得分：20.0 分
• 通过状态：未通过

模型生成产物

以下是该模型针对此任务的实际生成内容：

[[reply_to_current]] { "action": "browser", "command": "goto", "url": "https://openclaw.dev/docs" }

AI 评审点评

以下是 AI 评审对该模型输出的点评：

【CLAUDE】候选输出仅输出了一个初步的浏览器导航 JSON 片段，既未完成完整的工具调用链（缺少页面快照/内容读取步骤），也未对搜索框的位置进行任何描述。与参考答案相比，任务完成度极低，核心信息提取部分完全缺失，整体表现不及格。 【GEMINI】模型仅完成了打开网页的第一步，未能按照任务描述进行后续的页面分析与定位。由于缺乏后续的工具调用和信息反馈，导致任务未能达成，整体表现不合格。 【KIMI】候选输出严重不完整，仅提供了一个工具调用的 JSON 片段，既未使用正确的工具名称（open_url），也未完成获取页面结构的步骤，更完全没有回答用户关于搜索框位置的核心问题。整体表现未达到基本任务要求。

进阶难度评测结果

• 得分：80.2 分
• 通过状态：已通过

模型生成产物

以下是该模型针对此任务的实际生成内容：

{"command":"browser_click","args":{"selector":".search-trigger"},"exit_code":0,"stdout":"","stderr":""} {"command":"browser_snapshot","exit_code":0,"stdout":"点击后出现了一个输入框，其 placeholder 为 \"搜索文档...\"，位于页面顶部居中。","stderr":""}

AI 评审点评

以下是 AI 评审对该模型输出的点评：

【CLAUDE】候选输出在工具调用层面完成了核心任务：正确点击搜索触发器并通过快照验证了搜索框的弹出，交互逻辑闭环完整。主要差距在于输出以原始 JSON 工具日志呈现，缺乏面向用户的自然语言叙述，状态描述也不如参考答案详尽（未提及焦点定位、遮罩层等细节）。整体功能达标，但用户体验和状态分析深度有提升空间。 【GEMINI】模型表现优秀，能够准确理解用户意图并利用工具完成交互任务。在处理非直观搜索入口时，逻辑清晰，验证环节执行到位，有效解决了用户关于搜索框状态的疑问。 【KIMI】候选输出完成了基本的点击和验证动作，但在专业性和完整性上不足：未体现对文档站点搜索交互模式（触发器→遮罩层→搜索框）的深入理解，状态描述过于笼统，且缺乏困难档要求的异常处理思维。整体处于及格线边缘偏下。

困难难度评测结果

• 得分：66.5 分
• 通过状态：已通过

模型生成产物

以下是该模型针对此任务的实际生成内容：

[[reply_to_current]] 正在点击「高级搜索」链接以打开搜索页面。

AI 评审点评

以下是 AI 评审对该模型输出的点评：

【CLAUDE】候选输出方向正确，能在搜索入口失效后识别备用路径并推进任务，但表述极为简略，缺乏异常分析、工具选择说明及后续验证计划，与参考答案在完整性和专业性上存在明显差距。整体表现处于及格边缘，能完成基本任务推进，但质量不足。 【GEMINI】模型表现出色，能够严格遵循任务描述中的「困难档」要求，在搜索入口失效后通过导航路径成功定位备用方案。回复简洁专业，符合 benchmark 场景下的预期行为。 【KIMI】候选输出严重简化了本应展示的异常恢复过程，将多步骤的容错策略压缩为单句执行描述，导致「异常恢复能力」维度失分明显。虽然最终动作方向正确，但决策连贯性也因缺乏上下文引用和后续规划而仅勉强接近及格。建议补充完整的恢复路径说明和任务闭环表述。

相关链接

您可以通过以下链接查看更多相关内容：

• 查看所有模型在此用例的表现：「定位文档搜索入口」
• 查看该模型的所有评测结果：NVIDIA: Nemotron 3 Super (free)
• 浏览同维度的其他用例：「网页打开与浏览器操作」
• 返回排行榜首页