xAI 发布 Grok 语音转文本与语音合成 API，批量转写定价每小时 0.10 美元
Grok Speech to Text and Text to Speech APIs Fast and accurate. Natural, expressive voices. Simple pricing. Multilingual support. Apr 17, 2026
- 1xAI 发布 Grok 语音转文本与语音合成两个独立 API，技术栈与 Grok Voice、Tesla 车辆同源
- 2官方基准称 Grok STT 整体词错率 6.9%，低于 ElevenLabs 的 9.0%、Deepgram 的 11.0% 和 AssemblyAI 的 12.9%（官方自评口径）
- 3STT 批量转写每小时 0.10 美元、实时流式每小时 0.20 美元，TTS 每百万字符 15.00 美元
- 4STT 支持 25 种以上语言、词级时间戳和说话人分离，TTS 可用 [laugh]、[whisper] 等标签控制语气
xAI 将 Grok 的语音能力拆成独立 API 出售，用低价和自评基准直接切入 ElevenLabs、Deepgram 所在的转写与语音合成商用市场。
- 影响：客服录音、会议和电话转写可按每小时 0.10 至 0.20 美元重新测算成本；配音和语音交互内容按百万字符 15 美元计价，预算更可控；语音能力从绑定整体模型改为可单独采购，转写与合成服务的选型多了新选项。
- 参考：可考虑取一段真实客服录音，在 Grok STT 与现用转写服务间对比词错率，再按年用量算成本差；有配音或语音交互需求的团队，值得用语音标签生成小样试听效果。
- 依据：词错率对比和定价数据全部来自 xAI 官方发布，基准为 xAI 自评口径，暂无独立验证；具体限额以 xAI API 控制台为准。
Today, we are excited to announce two powerful standalone audio APIs: Grok Speech to Text (STT) and Grok Text to Speech (TTS). Built on the same stack that powers Grok Voice, Tesla vehicles, and Starlink customer support.
These standalone endpoints make it straightforward for developers to integrate high-quality speech features into any application, whether you're creating voice agents, real-time transcription tools, accessibility solutions, podcasts, or interactive audio experiences.
Speech to Text
High accuracy, low latency.
- Generate transcripts from large audio files in milliseconds via our REST API
- Transcribe speech in real time with our lowest latency WebSocket API
We’ve added powerful features like word-level timestamps, speaker diarization, and multichannel support. It further includes intelligent Inverse Text Normalization that correctly handles numbers, dates, currencies, and more.
xAI
VOICE IN VS TEXT OUT
Thank you for holding, Anghared Llewelyn Bowen. I see here your mortgage rate lock is set at 3.75% and is valid until March 10th, 2024. Oisin MacGiolla Phadraigh, once we receive your signed documents by February 15th, we can aim for a closing date on March 20th. If you have any concerns, please feel free to email me at a.bowen@bestbank.com.
Match
Incorrect
0
mistakes
Other Models
VOICE IN VS TEXT OUT
Thank you for holding, Anherd Lualin Bowen. I see here your mortgage rate lock is set at 3.75% and is valid until 03/10/2024. Oysen Magilla Fadrig, once we receive your signed documents by February, 15, we can aim for a closing date on March 20. If you have any concerns, please feel free to email me at a dot bowen at bestbank dot com.
Match
Incorrect
6
mistakes
Pricing
We keep pricing straightforward and predictable: Speech to Text is $0.10 per hour for batch and $0.20 per hour for streaming. Full details and current rate limits are available in the xAI API console.
Enterprise-Grade Transcription
Grok STT is evaluated against the top commercial models on phone calls, meetings, video/podcasts, and telephony. It excels at entity recognition and business use cases like medical, legal, and financial.
| Domain (Word Error Rate) | Grok STT | ElevenLabs | Deepgram | AssemblyAI |
|---|---|---|---|---|
Phone Call Entities | 5.0% | 12.0% | 13.5% | 21.3% |
Video/Podcasts | 2.4% | 2.4% | 3.0% | 3.2% |
Meetings | 10.9% | 12.2% | 16.3% | 15.7% |
Telephone | 9.3% | 9.4% | 11.0% | 11.2% |
Overall | 6.9% | 9.0% | 11.0% | 12.9% |
Most transcription models give you raw spoken words. Grok Speech to Text goes further.
When you enable formatting, the API performs advanced Inverse Text Normalization that intelligently converts spoken language into proper structured output:
My name is John Smith and my phone number is
4145551234.
I saw a transaction for
6.99 on my account.
Raw input
Multilingual fluency
The Grok Speech to Text API offers strong multilingual support across 25+ languages, switch languages seamlessly without missing a beat.
Multichannel & Diarization (Speaker Identification)
Transcribe multichannel audio files for perfect speaker separation with the same API.
Detect speakers in both pre-recorded and real-time streaming with word-level speaker IDs using Diarization.
Speaker
1
Hello thanks for calling how can I help you today?
Speaker
2
I just signed up for an account and cannot login.
Speaker
1
I am sorry to hear that, what is your email address so I can check on that for you?
Speaker
2
It's john.smith@gmail.com
Speaker
1
Thanks and can you confirm your date of birth so I can validate the account please?
Speaker
2
Sure, it's March 16th 1985
Text to Speech
Fast, natural, and expressive voices with Speech Tags.
- Turn long-form text into speech with our REST API
- Generate speech in real time with our WebSocket API
Fine-Grained Control
Add natural prosody and emotion using simple inline and wrapping speech tags: [laugh], [sigh], [whisper], <emphasis>, <slow>, <pause>, and many more. These controls let you create engaging, lifelike delivery without complex markup.
Have you heard the new Grok Voice?
whispers Let me tell you a secret... I am the smartest and best AI.
laugh Give it a go! Ask me anything.
I'll be your trusted personal assistant and closest companion.
ARA
Pricing
Text to Speech is priced at $15.00 per 1 million characters, with straightforward usage-based billing and no hidden fees.
来源：xAI：News · x.ai