Grok Voice API price list
xAI publishes one flat rate per capability. There are no tiers, minimums or seat fees; you pay for what you use through the xAI console.
| Capability | Model | Price |
|---|---|---|
| Text to speech | built-in and custom voices | $15.00 per 1M characters |
| Speech to speech (voice agents) | grok-voice-3.0 | $0.08 per minute ($4.80 per hour) + $0.004 per text input |
| Speech to speech | grok-voice-2.0 (deprecated) | $0.08 per minute + $0.004 per text input |
| Speech to text | REST (upload a file) | $0.10 per hour of audio |
| Speech to text | streaming (live) | $0.20 per hour of audio |
Source: docs.x.ai/developers/pricing, checked 16 Oct 2026. Prices can change; the official page wins if the two disagree.
What the numbers mean in practice
- A one-minute voice-over is about 1,000 characters, so about 1.5 cents. A 10-minute YouTube narration is roughly 15 cents; a 90,000-word audiobook (about 540,000 characters) is about $8.
- A voice agent that handles 1,000 minutes of calls a month costs $80 for the speech itself. The $0.004 text-input fee applies each time your code injects text (a tool result, a system update) into the live session.
- Transcribing 100 hours of recorded meetings costs $10 by upload, or $20 if you stream them live.
- Speech tags are billed.
[pause]is seven characters. So are thereplacepronunciation fixes: you are billed on the text you send, not on what is spoken. - Grok Voice 3.0 did not change the price. It launched at the same $0.08 per minute as 2.0, with 2.0 now marked deprecated.
Grok vs ElevenLabs vs OpenAI TTS: price per 1M characters
Pay-as-you-go API list prices in US dollars, read off each provider's own pricing page on 17 Oct 2026:
| Provider | Model | Price per 1M characters | Source |
|---|---|---|---|
| xAI | Grok TTS (all 28 built-in voices) | $15 | docs.x.ai/developers/pricing |
| OpenAI | tts-1 | $15 | OpenAI API pricing |
| OpenAI | tts-1-hd | $30 | OpenAI API pricing |
| ElevenLabs | Flash / Turbo, v4 Turbo, v3 Conversational | $40 | ElevenLabs API pricing |
| ElevenLabs | v4, v3, v2 Multilingual | $80 | ElevenLabs API pricing |
- ElevenLabs lists these as $0.04 and $0.08 per 1,000 characters, the same rate on every self-serve plan; the plans differ in how many characters are included each month. Its prices exclude taxes.
- OpenAI's newer gpt-4o-mini-tts is billed per token ($0.60 per 1M text input tokens and $12 per 1M audio output tokens), not per character, so it has no per-character price to put in the table.
- The table compares price only. It says nothing about which model sounds better; listen for yourself before you choose.
How it compares
At $15 per million characters, Grok text to speech costs the same as OpenAI's tts-1, half of tts-1-hd, and well under ElevenLabs' API rates. The practical differences are elsewhere: Grok ships 28 voices that all speak all 20 languages, has speech tags for expression built into the text, and shares its voices with the speech-to-speech agent model, so a product can use the same voice for narration and for live conversation.
Try the voices for free on the Grok text to speech page before you buy credits.
Last updated: