预约演示 免费试用 Speak AI 登录
说话 API

Transcription API plus the AI platform built on top of it.

一个账户可让您使用 API、MCP 服务器、CLI 和完整的工作区 UI。按量付费，无最低消费，信用额度永不过期，无需信用卡即可开始。

获取您的 API 密钥
查看按量付费定价
PAYG现收现付
100+语言
83MCP Tools
7-dayFree Trial
值得信赖 by 300,000+ people and teams

Most transcription APIs make you choose. You don’t have to.

选择转录 API 通常需要权衡。Deepgram 提供按小时按量付费，但没有 AI 层或顶层端用户 UI。AssemblyAI 提供 API 和 AI，但没有团队工作区。Otter 提供工作区，但将 API 限制在企业版后面，并使用订阅模式运营。Speak 是唯一一个在单一账户中提供全部四项功能的平台：转录 API、AI 平台、完整的网页和移动 UI，以及无最低消费、信用额度永不过期的按量付费定价。

按量付费。无需订阅。

Transcription is metered per second on a pay-as-you-go credit balance, and the AI Meeting Assistant runs at the same credit rate. AI Chat is billed in credits, based on usage. See the 定价页 有关当前费率的信息。随时按任意金额充值信用余额。信用额度永不过期。无合同、无月度最低消费，无需信用卡即可获取 API 密钥并开始测试。

Same rates whether you call the API, use the CLI, the MCP server, or upload through the dashboard. One balance, every surface.

查看完整按量付费定价 →

Three lines to your first transcript.

Authenticate, upload, get the transcript back. The API, CLI, and MCP server all hit the same endpoints and pull from the same credit balance.

Upload via REST API
curl https://api.speakai.co/v1/media/upload 
  -H "x-speakai-key: $SPEAK_API_KEY" 
  -H "Content-Type: application/json" 
  -d '{
    "name": "Interview",
    "url": "https://example.com/interview.mp3",
    "mediaType": "audio",
    "sourceLanguage": "en-US"
  }'

curl example. Full SDKs and reference at docs.speakai.co.

Connect any AI client in under a minute.

Speak MCP 服务器向任何支持 MCP 的 AI 客户端公开 83 个工具。根据您的团队和工具选择，有三种连接方式。使用相同的按量付费信用余额，无需单独计费。

One-Click Connect RECOMMENDED

Browser-based OAuth. Click Connect, approve the permissions popup, done. No API key to paste. Works for Claude.ai, ChatGPT, and any client that supports remote MCP with OAuth. Revoke any time at api.speakai.co/v1/oauth/connections.

API Key plus remote URL

Add the remote MCP endpoint https://api.speakai.co/v1/mcp to your client, then paste an API key generated at app.speakai.co/developers/apikeys. Paste-only setup, no local install. Best for Claude Desktop, Cursor, Windsurf, VS Code, Zed.

Developer CLI

Run npx @speakai/mcp-server init in your terminal. Auto-detects your Claude Desktop, Claude Code, or Cursor install and wires it up. Best for self-hosters and engineers who want the server running locally.

Works with the AI clients your team already uses:

Claude.aiWeb
Claude DesktopMac, Win
Claude Code终端
ChatGPTMCP connector
光标IDE
VS CodeCopilot
WindsurfIDE
ZedEditor

83 tools across 10 categories. Everything your AI client can do with your Speak workspace:

媒体库

Upload, list, search, tag, favorite, move, delete media across your workspace.

Magic Prompt

Run custom AI prompts across one file or your entire library. Score, classify, extract.

搜索

Full-text search across every transcript, with filters by tag, folder, date, speaker.

文件夹

Create folders, move media, manage shared library structure for your team.

Recorders

Create branded recorder URLs, send to respondents, pull in submissions automatically.

Clips

Extract verbatim clips from transcripts, export with captions, share or embed.

会议机器人

Schedule the AI Meeting Assistant for Zoom, Teams, Meet, Webex, and Outlook events.

自动化

List, create, toggle, and update workspace automations that run on every new media.

Text notes

Create and update text-only documents alongside audio and video transcripts.

嵌入播放器

Generate public or private embed URLs for transcripts, clips, and analyses.

Prefer no code at all? The Speak Zapier app exposes 5,000+ integrations. Trigger on a new Speak transcript, action on Slack, Notion, HubSpot, Salesforce, Google Sheets, and more. Browse the Speak Zapier app →

获取您的 API 密钥
Full MCP docs

How Speak compares.

2026 年的按量付费转录 API 市场有三个主要参与者：AssemblyAI、Deepgram 和 Speak。Otter 涉足工作区类别。Speak 是唯一一个在单一账户中提供全部四个层级（API + AI + 工作区 UI + 按量付费）的平台。

Speak AI

All four layers

  • ✓Transcription API
  • ✓按量付费定价
  • ✓AI Chat over transcripts
  • ✓End-user workspace UI
  • ✓MCP server (83 tools)

AssemblyAI

API plus AI, no workspace

  • ✓Transcription API
  • ✓按量付费定价
  • ✓AI layer (LeMUR / Claude)
  • ✗End-user workspace UI
  • ✗MCP server

Deepgram

API plus PAYG, no AI or UI

  • ✓Transcription API
  • ✓按量付费定价
  • ✗AI Chat / analysis layer
  • ✗End-user workspace UI
  • ✗MCP server

水獭

Workspace UI, no PAYG API

  • ✗API (Enterprise only)
  • ✗按量付费（仅限订阅）
  • ~AI layer (limited)
  • ✓End-user workspace UI
  • ✗MCP server

Speak 是唯一将接近 API 的定价与完整 AI 工作区配对的平台。AssemblyAI 提供 API。Otter 提供 UI。Deepgram 提供按量付费。Speak 提供所有三项，加上 AI Chat、MCP 和 CLI。

Everything in one account.

One API key, one credit balance. Use any surface, any time.

Transcription API

100+ 种语言、说话人分化和字级时间戳。REST 加 webhooks。按量付费信用余额的按秒计费。

AI Chat over your transcripts

Ask questions across one file or your entire library. Usage-based pricing with model routing handled for you. No separate AI subscription required.

MCP server

83 tools for Claude, Cursor, Windsurf, and any MCP-aware client. Install with npm install and you’re done. Same pay-as-you-go credit balance.

CLI

26 commands to transcribe, search, export, and analyze from your terminal. MIT licensed. Works with the same API key and the same credit balance.

Built for usage that flexes.

No credit card to start.

Sign up, get an API key, run your first call. Pay only when you add credits. The 7-day trial includes full API access and the MCP server at no charge.

Credits never expire.

充值 $10 或 $1,000。按量付费余额在您的账户中保留，直到您使用它。无月度消耗。无过期日期。六个月后回来，您的信用额度仍然存在。

No contracts, no minimums.

通过停止使用即可取消。无合同、无月度最低消费，无锁定。即使您停止使用该产品，您的按量付费信用余额也会保留在您的账户中。

Teams that build on Speak AI

★★★★★
4.9 G2

“Speak AI 在改变我们处理定性数据的方式方面发挥了重要作用。转录准确度令人印象深刻，NLP 见解为我们节省了数小时的手动分析。”

研究总监 | 咨询公司

“我们从 Otter.ai 切换过来，分析深度完全不同。情感评分、关键词提取和主题检测全部自动进行。”

产品经理 | SaaS 公司

“能够在所有客户通话中搜索并提取特定时刻，这对我们的支持团队来说改变了游戏规则。”

客户体验负责人 | 企业技术

Who uses the Speak API

一个 API 密钥，一个按量付费信用余额，涵盖在语音和音频数据上构建的每个团队。

Product engineers

Build transcription directly into your product.
Upload audio via REST, subscribe to the media.analyzed webhook, and get back a structured transcript with speaker labels, word-level timestamps, topics, sentiment, and entities.
按量付费计费意味着您只需为处理的内容付费。
No subscription tier required, no monthly minimum.

Data and analytics teams

Run AI Chat over your entire transcript library.
Ask questions across thousands of calls, interviews, or recordings.
Billed in credits, based on usage, on the same pay-as-you-go credit balance.
Use Magic Prompts to extract structured data at scale: MEDDIC scores, objection flags, competitor mentions, NPS signals.

AI engineers and agents

Connect the Speak MCP server to Claude, Cursor, Windsurf, or any MCP-aware client.
83 tools cover transcription, search, export, clip creation, folder management, and more.
Your AI agent can upload audio, wait for analysis, and return structured answers, all through conversation.
与 REST API 相同的按量付费信用余额。

销售和 RevOps 团队

Auto-transcribe every call your dialer records.
Feed structured data into your CRM on the media.analyzed webhook.
Run Magic Prompts for discovery scoring, competitor extraction, and follow-up action items.
无需更改订阅层级。按量付费信用额度随您的调用量扩展。

How a transcript goes from upload to insights

四个步骤。全部采用同一个按量付费信用余额。

上传

POST your audio or video file URL to https://api.speakai.co/v1/media/upload with your API key in the x-speakai-key header.
Speak accepts public URLs, Twilio recording URLs with embedded auth, and direct file bytes.
Specify source language or let Speak detect it automatically.

转录

Speak queues and processes the recording.
Transcription includes speaker diarization, word-level timestamps, and support for 100+ languages.
A 10-minute call typically completes within 60 seconds.
按秒计费，扣费来自您的按量付费信用余额。

分析

Once the transcript is ready, Speak runs AI analysis automatically: topics, sentiment, named entities, keywords, and any Magic Prompts you have saved for this folder.
The same credit balance covers transcription and analysis.
No separate AI billing required.

React

Speak fires the media.analyzed webhook to your registered endpoint when everything is ready.
Fetch the full transcript and insights from GET /v1/media/insight/:mediaId.
Push structured data to your CRM, trigger downstream automation, or surface insights in your own UI.

获取您的 API 密钥
查看 API 文档

Transcription API with AI platform: why they belong together

大多数转录 API 都是点对点解决方案。您获得一个端点、一个响应和一份转录文本。您如何处理该转录文本完全是您自己的问题。您需要单独的 LLM 调用来进行摘要、单独的数据库来存储和搜索转录、单独的 UI 供非技术团队成员使用，以及为添加的每一层单独的计费关系。Speak 的构建基于这样的前提：这些层应该在一个账户中汇聚，采用同一个按量付费信用余额，无需订阅即可开始使用。

按量付费对 API 开发者意味着什么

Pay-as-you-go on Speak means a prepaid credit balance that you top up in any amount. There is no monthly subscription charge, no minimum monthly spend, and no expiration date on credits you add. Transcription is metered per second so a 4-minute call costs roughly the same fraction of an hour as it takes. The AI Meeting Assistant and the transcription API share the same credit rate. AI Chat over your transcripts is billed in credits, based on usage. All of this pulls from one credit balance regardless of which surface you use: the REST API, the CLI, the MCP server, or the dashboard.

For developers this means you can onboard without a procurement cycle, test a full pipeline without committing to a subscription tier, and scale usage up or down without renegotiating. If you stop using Speak for two months, your credit balance is waiting for you when you come back.

How the API fits into the broader platform

Speak REST API 是基础。产品中的其他每个界面都基于相同的端点构建。CLI 用 26 个终端命令包装了上传、搜索、导出和分析端点。MCP 服务器向 Claude、Cursor、Windsurf 和任何兼容 MCP 的 AI 客户端公开 83 个工具。Web 和移动 UI 让非技术团队成员使用与您通过 API 触发的相同转录和 AI 分析功能。一个 Speak 账户涵盖所有四个界面。一个按量付费信用余额覆盖所有使用。

Transcription API plus AI Chat: the compound value

Transcription gets you a text representation of audio. AI Chat over that transcript gets you answers. Ask a question across one file or your entire library. Identify speakers, extract decisions, find every call where a competitor was mentioned, or summarize the 50 longest interviews in your corpus. AI Chat is billed in credits, based on usage, model routing is handled for you, and the same pay-as-you-go credit balance covers it. No second account, no second API key, no second billing relationship.

MCP server for Claude and other AI clients

Speak MCP 服务器让 AI 助手能够直接访问您的 Speak 工作区。Claude Desktop、Claude Code、Cursor 和 Windsurf 可以转录新音频、搜索现有转录、运行 Magic Prompt、提取演讲者片段、导出内容等，所有这些都通过自然语言完成。使用 npm 安装 MCP 服务器，粘贴您的 Speak API 密钥，任何兼容 MCP 的客户端都有 83 个工具可用。相同的按量付费信用余额覆盖所有 MCP 触发的使用。无需额外的计费配置。

Getting started with the Speak transcription API

Sign up for a free account at app.speakai.co. Open the Developers section of your dashboard to generate an API key. No credit card required. The 7-day trial includes full API access. When you are ready to scale, top up your pay-as-you-go credit balance in any amount. Credits never expire. Full API reference, SDK guides, and webhook documentation are at docs.speakai.co.

Webhooks and async analysis

Speak fires a signed webhook when transcription and AI analysis are complete. Register your endpoint via POST /v1/webhook and specify which event types you want. The media.analyzed 当转录和所有 AI 见解准备就绪时触发的事件。使用它来触发下游自动化：将转录推送到您的 CRM、发送 Slack 警报、更新数据库行，或在每个新录音上运行自定义 Magic Prompt。Webhook 是同一按量付费账户的一部分，使用相同的 API 密钥，相同的信用余额。

Speaker diarization and multi-language support

The Speak transcription API supports 100+ languages and automatic language detection. You can specify sourceLanguage on upload or let Speak identify the language for you. Speaker diarization labels each turn in a conversation by speaker, making transcripts of interviews, calls, and meetings immediately parseable without manual cleanup. Word-level timestamps are available on every transcript so you can build video chaptering, subtitle files, or searchable moment-in-time interfaces on top of the API response.

2026 年按量付费 API 提供商对比

2026 年的按量付费转录 API 市场有三个有意义的选项：AssemblyAI、Deepgram 和 Speak。如果您需要仅限 API 的工作流且具有内置 LLM 工具（LeMUR）但没有终端用户工作区，AssemblyAI 是最强的选择。Deepgram 具有最低的原始转录小时费率和强大的实时流媒体支持，但没有 AI 分析层或工作区 UI。当您需要转录加 AI Chat 加工作区 UI（非技术团队成员可以使用），全部在一个按量付费信用余额下且无需订阅，Speak 是正确的选择。2026 年的空白象限属于 Speak：唯一在单个账户中提供所有四层的供应商。

Using the Speak CLI and MCP server from your API key

Your Speak API key works across every surface. The CLI lets you transcribe a local file with one command: speak transcribe interview.mp3。相同的 API 密钥和相同的按量付费信用余额。MCP 服务器对 Claude 和 Cursor 访问使用相同的密钥。您永远不需要管理多个密钥、多个计费账户或多个凭证集。一个 Speak 账户、一个密钥、一个余额、每个界面。

Questions developers ask.

I don’t write code. Is the API for me?

Probably not. If you want to upload media, run AI Chat over your transcripts, share folders with your team, or use the Meeting Assistant, start at speakai.co/pricing/ and pick a workspace plan. The API is for engineers integrating Speak into their own software.

Do I need a subscription to use the API?

否。API 从第一天起就按量付费运行。随时添加信用，金额任意。订阅存在于希望获得包含时数和团队功能的团队，但调用 API 永远不需要订阅。按量付费信用余额可以独立运行，无需附加订阅。

How do I get API keys?

Sign up for a free account, open the Developers page in your dashboard, and generate a key. No card required. Everyone with a Speak account can generate API keys.

Which AI clients work with Speak?

Eight today: Claude.ai (web), Claude Desktop, Claude Code, ChatGPT, Cursor, VS Code, Windsurf, and Zed. Anything that supports MCP works. Use One-Click Connect for Claude.ai and ChatGPT, paste the remote MCP URL plus an API key for desktop apps and IDEs, or run npx @speakai/mcp-server init 用于本地设置。相同的 83 个工具和相同的按量付费信用余额在每个客户端上。

What is the rate?

Transcription and the AI Meeting Assistant are both metered per second on a pay-as-you-go credit balance, and AI Chat is billed in credits, based on usage. The same rates apply whether you use the API, the CLI, the MCP server, or upload through the dashboard. Full pricing at speakai.co/pricing/.

Get your API key in 60 seconds.

No card. No call with sales. No subscription. Sign up, generate a key, make your first call. Pay only when you add credits.

获取您的 API 密钥

Create a free account, open the Developers page, and generate an API key. Full pay-as-you-go access from day one. 7-day trial included. No credit card required.

获取您的 API 密钥
登录

Explore the docs

上传端点、webhook 事件、Magic Prompt API、MCP 服务器工具和 CLI 命令的完整参考。从按量付费信用余额开始，然后从那里扩展。

查看 API 文档
按量付费定价
所有集成
MCP
CLI
开发者
定价
Docs