Otter.aiOtter.ai
Otter 是「机器人入会」的会议转写纪要产品:实时转写+行动项+跨会议问答,英文向成熟; 机器人进你的会议这件事本身要过合规。
Otter is the bot-joins-the-meeting transcription & notes product: real-time transcription + action items + cross-meeting Q&A, mature for English; the bot entering your meeting is itself a compliance question.
适合与不适合When to use it, when not to
适合:英文会议/访谈记录、小团队纪要流水线。 不适合:中文会议主打、隐私极敏感场景、免费档当长期方案。
Fits: English meetings/interviews, small-team note pipelines. Not a fit: Chinese meetings as the main use, privacy-critical settings, treating the free tier as a long-term plan.
中文生态China availability 国产 / 国内可直接用 / 备案状态 · 快照domestic / direct access / CAC filing · snapshot 2026-10-22
- 出品属地Origin
- 国外出品foreign-made Otter.ai(美国旧金山,原 Blue Note 公司)。Otter.ai (San Francisco, US; formerly Blue Note).
- 国内使用Direct access
- 大陆可直连direct OK otter.ai 大陆直连实测 200(2026-10-22)。otter.ai measured 200 direct from mainland (2026-10-22).
- 备案状态CAC filing
- 未见于备案名单not in filing list:未见于网信办公示名单(2026-10-22 检索);该产品未面向大陆公众提供服务,备案不适用。not in the CAC public registry (searched 2026-10-22); the product does not serve the mainland public, so filing does not apply.
备案信息以网信办《生成式人工智能服务已备案信息》公示名单为准(按批次滚动更新,本站只抄编号不解读);「大陆可直连」是本站从大陆网络直连实测的结果,不是厂商承诺。GitHub 类站点当日实测有波动,口径写在依据里。Filing status is quoted from the CAC public registry (updated batch by batch; this site copies the number and does not interpret it). "Direct access" is this site's own reachability measurement from a mainland network, not a vendor promise. GitHub-hosted sites fluctuated on the measurement day — the note says so.
固定坐标系Fixed coordinate system
8 个维度dimensions, 与同分区其他对象逐项可比each comparable item by item against others in the same section.
- 模型与开放条件Model & openness
自研转写/纪要模型线(Otter 自述转写引擎);用户不选模型。 曾有「用客户音频训练」的隐私争议史(2021-2022 公开报道叙事)——当前训练口径按官方 隐私页(未逐条核验)。
A self-developed transcription/minutes model line (Otter's self-described transcription engine); users do not pick models. There was a privacy controversy over "training on customer audio" (2021-2022 public-report narrative) — the current training stance follows the official privacy page (not verified item by item).
- 运行位置Where it runs
网页+移动 App+会议平台机器人接入(Zoom/Meet/Teams 机器人入会实时转写,官方产品面); 处理在厂商云端。
Web + mobile App + meeting-platform bot joining (Zoom/Meet/Teams bots join meetings for live transcription, official product surface); processing on the vendor's cloud.
- 本地文件Local files
会议录音/转写/纪要存账户侧云端;导出(txt/csv/srt 等)归本地。未见官方声明的本机处理形态。
Meeting recordings/transcripts/minutes stored on the account-side cloud; exports (txt/csv/srt etc.) count as local. No officially stated local-processing form seen.
- 关机后的任务Tasks after shutdown
实时入会转写是核心形态(机器人常驻会议);批处理上传并存;无「关机续跑」概念(云服务)。
Live bot-joined meeting transcription is the core form (a bot resides in the meeting); batch upload alongside; no "keep-running-after-shutdown" concept (cloud service).
- 工具与扩展Tools & extensions
实时转写+说话人分离+自动纪要/行动项(官方产品面)、Otter Chat(对转写提问,官方产品面)、 协作编辑与分享;企业面有管理控制台(官网所见)。
Live transcription + speaker diarization + auto minutes/action items (official product surface), Otter Chat (ask questions of the transcript, official product surface), collaborative editing and sharing; the enterprise side has an admin console (seen on the official site).
- 上下文与记忆Context & memory
跨会议知识库是产品叙事(历史会议可检索/问答)——这是它的「记忆」面,细节未逐条核验。
A cross-meeting knowledge base is the product narrative (past meetings searchable/answerable) — that is its "memory" surface, details not verified item by item.
- 权限与限制Permissions & limits
账号+订阅+企业体系(D9:价格不采);免费档有月度转写分钟上限(官方定价页叙事,具体额度未逐条核验); 会议录音涉多方隐私——机器人入会的知情同意是使用方责任(条款口径未逐条核验)。
Account + subscription + enterprise system (D9: pricing not collected); the free tier has a monthly transcription-minutes cap (official pricing-page narrative, exact quota not verified item by item); meeting recordings involve multiple parties' privacy — informed consent for a bot joining is the user's responsibility (terms stance not verified item by item).
- 适合什么任务Best for
适合:会议记录/访谈整理(英文向成熟)、个人与小团队的转写纪要流水线。 不适合:中文会议主打(转写引擎英文向叙事)、隐私极敏感场景(云端+机器人入会)、 把免费档当长期方案(分钟数限制)。
Fits: meeting/interview documentation (mature for English), transcription-minutes pipelines for individuals and small teams. Does not fit: Chinese meetings as the mainstay (the transcription engine has an English-leaning narrative), extremely privacy-sensitive scenarios (cloud + bot joining), treating the free tier as a long-term plan (minute limits).
头号误解Top misconception
- 机器人入会模式=第三方机器人进你的会议——企业使用前要过 IT/合规审查,知情同意是使用方责任Bot-joins mode = a third-party bot entering your meeting — enterprises need IT/compliance review; informed consent is the user's responsibility
- 「用客户音频训练」争议史(2021-2022 公开报道)——当前口径以官方隐私页为准,敏感场景读条款原文A public 'trained on customer audio' controversy (2021-2022 reporting) — current practice follows the official privacy page; sensitive settings should read the terms text
- 转写引擎英文向成熟——中文会议准确率要实测,别按英文体验推断The engine is English-mature — Chinese meeting accuracy needs testing; don't infer from the English experience
未知项清单Known unknowns
- 当前训练口径(隐私页未逐条核验——争议史后的政策走向要读原文)
- 免费档分钟额度的当前值(未逐条核验)
- 中文转写模型线(未核验到官方声明)
- The current training stance (privacy page not verified item by item — the policy direction after the controversy requires reading the original text)
- The free tier's current minute quota (not verified item by item)
- A Chinese transcription model line (no official statement verified)
证据来源Evidence sources
判断可回到以下一手源复核。本站核验日Every judgement can be rechecked against the primary sources below. Verified on 2026-10-22,内容更新日content updated on 2026-10-22。
| 类型Type | 名称Name | 链接Link |
|---|---|---|
| official | Otter · 官网(产品与能力叙事)Otter · official site (product and capability narrative) | https://otter.ai |
| official | Otter · Terms of Service(官方条款,2026-10-22 直抓核验)Otter · Terms of Service (official terms, direct-fetch verified 2026-10-22) | https://otter.ai/terms-of-service |
| changelog | Otter · Blog(官方更新线,changelog)Otter · Blog (official update line, changelog) | https://otter.ai/blog |
实测记录Measured runs
本档未跑音频质量与延迟测试;结论只覆盖能力面与许可面。
实测建议(三条,都指向官方可读页没给的东西):
- 中文/中英混说会议的转写准确率
- Otter Chat 对多场历史会议跨检索的召回质量
- 机器人入会在企业 Zoom 策略限制下的实际可达性
This site has not completed measured runs. The coding-oriented test protocol is at protocols/coding/_protocol.md (the coding task suite does not apply to creation products).
Suggested measurements (three, all pointing at things the official readable pages do not give):
- Transcription accuracy on Chinese / Chinese-English mixed meetings
- Otter Chat's recall quality across multiple past meetings
- Bot joining's actual reachability under enterprise Zoom policy restrictions
相关条目Related entries
- DeepgramDeepgram上游对照:Otter 类产品的转写底座可以是 Deepgram/Whisper 这类引擎。Upstream contrast: the transcription base of Otter-class products can be engines like Deepgram/Whisper.
- AssemblyAIAssemblyAI同为转写+纪要向:AssemblyAI 给开发者 API,Otter 直接给终端用户产品。Also transcription + minutes oriented: AssemblyAI gives developers an API, Otter gives end users a product directly.
- WhisperWhisper开源基座对照:不想让会议音频出本机,就用 Whisper 自建管线。Open-source-base contrast: if you don't want meeting audio leaving your machine, build your own pipeline on Whisper.
- ZapierZapier纪要下游:Otter 导出的纪要常经 Zapier 类平台分发到任务/日历(建档后互链)。Minutes downstream: Otter's exported minutes are often distributed to tasks/calendars via Zapier-class platforms (cross-linked after filing).
本页由This page is generated from the ai-handbook 数据层生成(CC BY 4.0)。data layer (CC BY 4.0).
方法论与坐标系定义见仓库内For the methodology and coordinate definitions, see METHODOLOGY.md.