--- title: "Grok Voice Agent API: Pricing, Setup, Limits" description: "Grok Voice Agent API explained: how a real-time session works, the $0.08/min price, how it stacks up against OpenAI Realtime, and a pre-launch checklist." url: https://skipcall.io/en/blog/grok-voice-agent-api-guide language: en last_updated: 2026-11-03 --- [Ask Claude about this article](https://claude.ai/new?q=I'm%20reading%20%22Grok%20Voice%20Agent%20API%3A%20How%20It%20Works%20and%20What%20It%20Costs%22%20\(https%3A%2F%2Fskipcall.io%2Fen%2Fblog%2Fgrok-voice-agent-api-guide\).%20Read%20the%20article%2C%20then%20answer%20my%20questions%20about%20it.) [Ask ChatGPT about this article](https://chatgpt.com/?q=I'm%20reading%20%22Grok%20Voice%20Agent%20API%3A%20How%20It%20Works%20and%20What%20It%20Costs%22%20\(https%3A%2F%2Fskipcall.io%2Fen%2Fblog%2Fgrok-voice-agent-api-guide\).%20Read%20the%20article%2C%20then%20answer%20my%20questions%20about%20it.) LinkedIn X [Back to blog](https://skipcall.io/en/blog) Voice AI Published 3 November 2026 # Grok Voice Agent API: How It Works and What It Costs xAI's Grok Voice Agent API lets developers put Grok's real-time voice behind an app, a help line or a booking agent. Here is how a session actually works, what a minute costs in November 2026, how it compares with OpenAI's Realtime API, and the questions to settle before you ship it to production. $4.80 per hour of conversation with Grok Voice 3.0, the same rate as Voice 2.0, before tools and telephony < 1 s response latency xAI claims for Grok Voice 3.0, released October 14, 2026 $240 a month for 1,000 three-minute calls at today's rate, before tools, phone numbers and carrier minutes The Grok Voice Agent API is the developer layer of xAI’s voice stack: you stream a caller’s audio to Grok over a WebSocket, and Grok streams a spoken reply back in under a second, according to xAI. It costs $0.08 per minute in November 2026, and it is the building block behind custom help lines, booking agents and in-app voice assistants. If you are deciding whether to build on it, this guide covers what matters in practice: how a session flows, what you will actually pay, where it differs from OpenAI’s Realtime API, and what to clear with legal and security before go-live. For the big picture of Grok Bot, the API and the no-code Builder, start with our [guide to Grok and phone calls](https://skipcall.io/en/blog/grok-phone-calls-guide). 40% less latency than GPT-Live-1 on the τ-Voice benchmark, according to xAI's Grok Voice 3.0 announcement 3 launch voices, Ara, Eve and Leo, with tags to whisper, sigh or laugh mid-sentence Dec 17, 2025 launch date, at $0.05 per minute, with an official plugin for LiveKit ## What does the Grok Voice Agent API actually do? Think of it as a live phone line between your software and Grok. Your application sends audio, from a browser microphone or a phone call, and Grok listens, decides what to say and answers out loud. The connection stays open the whole time, so the conversation flows instead of advancing one request at a time. xAI shipped it on December 17, 2025, priced at $0.05 per minute, with an official LiveKit plugin. LiveKit is an open-source platform many teams already use to build voice agents, so the plugin gave them a short path to production. xAI also describes the API as compatible with the OpenAI Realtime specification, the de facto format for real-time voice. In xAI’s lineup, the API sits in the middle. Above it are two finished products: [Grok Bot, which can place calls for you](https://skipcall.io/en/blog/grok-bot-phone-calls), and the [Grok Voice Agent Builder](https://skipcall.io/en/blog/grok-voice-agent-builder-review), which creates a phone agent without code. Below it are the voice models themselves, most recently Grok Voice 3.0. ## Why does a voice-to-voice model feel faster? Older voice assistants chained three systems: speech-to-text, a language model that wrote a reply, and text-to-speech that read it out. Each hop added delay, and tone of voice got lost because the model only ever saw words. LiveKit describes the xAI model as voice-to-voice: sound goes in and sound comes out of one model. In practice that gives you two things: - **Shorter pauses.** LiveKit reports replies in under 700 milliseconds. xAI’s figure for Voice 3.0 is under one second, and it claims 40% lower latency than GPT-Live-1 on the τ-Voice benchmark. - **More expressive delivery.** The model can laugh, sigh or whisper, and xAI exposes that through tags such as \[laugh], \[sigh] and \[whisper]. We unpack the voice side, from barge-in to the 8 kHz sound of a phone line, in [how Grok talks on the phone](https://skipcall.io/en/blog/how-grok-talks-on-the-phone). > Callers forgive a slightly odd word choice. They do not forgive a two-second silence after every sentence. ## How does a session with the API work, step by step? Exact event names depend on your SDK and version, so treat this as the shape of every session rather than copy-paste code. The transport is a WebSocket, a channel that stays open for the whole conversation instead of firing one request per message. ![Diagram of a Grok Voice Agent API session in five steps between your application and Grok's real-time API: 1 open the session over a persistent WebSocket, 2 settings for voice, instructions and allowed tools, 3 the caller's audio streams to Grok, 4 Grok answers in audio and can call a tool, 5 the session ends and is billed per minute at $0.08; notes say the only region listed in the docs is us-east-1 and the API is compatible with the OpenAI Realtime spec](https://skipcall.io/assets/infographies/grok/figures/en/session-api.jpg) Five steps, one open connection: the meter runs per minute of conversation, so a session your code forgets to close keeps costing money. 01 #### Open the session per conversation Your server opens an authenticated WebSocket to xAI with your API key. It stays open until the call ends. 02 #### Send the settings milliseconds You send the agent’s instructions (role, tone, goal), the voice, the audio format and the tools it may call, such as an availability lookup in your calendar. 03 #### Stream the caller's audio whole call Audio goes up in small chunks. For phone calls, xAI’s official example keeps the telephone format, μ-law at 8 kHz, with no conversion. 04 #### Play Grok's answer and run tools real time Grok streams audio back along with events: speech started, speech stopped, transcript, tool requested. Your code runs the tool and returns the result to the model. 05 #### Close and log end of call Close the socket, then store what you need (transcript, summary, duration) according to your disclosure and retention policy. Session length drives the bill, at $0.08 per minute. To put this behind a real phone number, xAI publishes an official example that connects a Twilio number to the API. Numbers, SIP trunks and porting are covered in our guide to [connecting Grok to a phone number](https://skipcall.io/en/blog/connect-grok-to-a-phone-number). ## How much does the Grok Voice Agent API cost? You pay per minute of conversation, and the rate has changed since launch. Many blog posts still quote the old number. | Period | Model | Price per minute | Per hour | | - | - | - | - | | December 2025 (launch) | grok-voice-think-fast-1.0 | $0.05 | $3.00 | | From August 5, 2026 | Grok Voice 2.0 | $0.08 | $4.80 | | From October 14, 2026 | Grok Voice 3.0 | $0.08 | $4.80 | Three more lines usually show up on the invoice: - **Tools.** Web search and other tool calls are billed separately. Trade press reports $5 per 1,000 web searches. - **Telephony.** If the agent answers the phone, your carrier, for example Twilio, bills the number and its own per-minute rate. - **Hosting.** The small server that bridges the phone call and the API runs on your infrastructure. A quick sizing example: a help line handling 1,000 calls a month at an average of three minutes uses 3,000 minutes, which is $240 of Grok conversation before tools, numbers and carrier minutes. At 10,000 calls, it is $2,400, and the carrier side starts to weigh as much as the AI side. ## Grok Voice Agent API vs. OpenAI Realtime: what is actually different? Because xAI claims spec compatibility, the comparison comes up in every evaluation. Here is what the sources support, without guesswork. | | Grok Voice Agent API | OpenAI Realtime API | | - | - | - | | **Generally available** | December 17, 2025 | August 28, 2025 (gpt-realtime) | | **Model type** | Voice-to-voice, per LiveKit | Real-time model gpt-realtime | | **Stated latency** | Under one second (xAI) | Not compared here | | **Wire format** | Compatible with the OpenAI Realtime spec | The reference spec | | **Phone calls** | Official example through Twilio Media Streams | SIP calling supported | | **Extras cited** | \[laugh], \[sigh], \[whisper] tags, LiveKit plugin | MCP servers, image input | | **Price** | $0.08/min (Voice 2.0 and 3.0) | Not compared here | | **Region listed in docs** | us-east-1 | Not compared here | “Not compared here” means we did not verify the figure, not that it does not exist. The practical difference for a phone project is the plumbing. OpenAI documents native SIP calling, while xAI’s official phone example bridges through Twilio Media Streams. Both work, but they are set up differently, and compatibility at the message level does not mean identical behavior. Run the same call scripts on both before you commit. ## What should you settle before going to production? Ask xAI these five questions in writing and keep the answers with your vendor file: 1. **Where are audio and transcripts processed and stored?** Is any region other than us-east-1 available? 2. **How long is data retained, and can you request deletion?** 3. **Are conversations used to train models?** If so, can you opt out? 4. **Which subprocessors touch the data?** Hosting, telephony, any tool providers. 5. **What transfer mechanism covers non-US callers?** Ask which one applies to your contract. Then check the calling rules that apply to your use case: - **Outbound calls in the US.** In February 2024, the FCC confirmed that AI-generated voices are “artificial” voices under the TCPA, so consent rules for artificial-voice calls apply. The FCC’s consent revocation rules changed again on November 25, 2026, as we explain in [the TCPA changes for B2B teams](https://skipcall.io/en/blog/tcpa-rule-changes-november-2026). Our [cold calling compliance checker](https://skipcall.io/en/tools/cold-calling-compliance-checker) walks through the basics. - **Callers in the EU.** Since August 2, 2026, Article 50 of the AI Act requires telling people they are talking to an AI, at the latest at the first interaction. - **Recording notices.** If you record or transcribe, say so at the start of the call. Our [call recording notice generator](https://skipcall.io/en/tools/call-recording-notice-generator) drafts the wording. ## Who should build on the API, and who should not? **Software vendors** Teams adding a natural voice to their own product, with engineers who can own the integration and its upgrades. **Inbound phone lines** Reception, triage, scheduling and FAQs, with a clean handoff to a person. Inbound is where voice agents earn their keep today. **Not for a quick test** No developer, or just validating the idea? The no-code Builder, or a well-run phone system, will save you weeks. One reality check for sales leaders: a voice agent does not replace a rep on a prospecting call. Most revenue teams still win with good business numbers and tools that make humans faster. That is Skipcall’s job: a VoIP phone system for sales reps with a power dialer, transcription and CRM sync, from $25 per month (billed annually), no commitment. ## The bottom line The Grok Voice Agent API gives developers xAI’s real-time voice over a WebSocket, compatible with the OpenAI Realtime spec, at $0.08 per minute with Grok Voice 3.0. It is fast and expressive, but young: prices changed twice in ten months, the docs list only us-east-1, and outbound use in the US runs into TCPA rules on artificial voices. Build on it when you have engineers and a clear inbound use case, and test it against the alternatives first. For how the API fits next to Grok Bot and the Builder, head back to our [Grok phone calls guide](https://skipcall.io/en/blog/grok-phone-calls-guide). _Sources: xAI, “Grok Voice Agent API” (December 17, 2025) and “Grok Voice 3” (October 14, 2026); xAI real-time API documentation; LiveKit, xAI partnership announcement (December 17, 2025); GitHub xai-org/xai-cookbook, voice-agent-phone example; InfoQ on gpt-realtime general availability (September 2025); FCC Declaratory Ruling FCC 24-17 (February 8, 2024); Regulation (EU) 2024/1689, Article 50._ Author [Charles Baldet](https://skipcall.io/en/author/charles-baldet) [Charles Baldet on LinkedIn](https://www.linkedin.com/in/charlesbaldet/) CEO & Co-Founder, Skipcall Charles is the CEO and co-founder of Skipcall. A sales commando with over 10 years of experience in B2B SaaS and complex strategic accounts, he has closed major deals with Stellantis, SNCF, RATP and Natixis. A specialist in the PUCCKA and MEDDIC methodologies, Charles regularly teaches sales at HEC's incubator and the Sorbonne. He was ranked among Les Echos' top 10 business angels under 35 in 2020. He also co-founded Getalead (B2B sales agency) and Getlab (SalesTech studio). [See all their articles](https://skipcall.io/en/author/charles-baldet) ## A calendar full of meetings starts with a call. [Get started now](https://app.skipcall.io/signup) [Book a demo](https://skipcall.io/en/demo) **No commitment** · Number live in 2 minutes FAQ ## Frequently asked questions **What is the Grok Voice Agent API in one sentence?** It is xAI's developer interface for holding a live spoken conversation with Grok: your code streams audio in over a WebSocket, and Grok streams its spoken answer back, for as long as the conversation lasts. **How much does the Grok Voice Agent API cost per minute?** With Grok Voice 3.0, released October 14, 2026, it costs $0.08 per minute of conversation, or $4.80 per hour, the same price as Voice 2.0. At launch in December 2025 the rate was $0.05 per minute. Tools the agent calls, such as web search, are billed on top, and your phone carrier bills its own minutes separately. **Can I reuse code written for the OpenAI Realtime API?** Partly. xAI describes its API as compatible with the OpenAI Realtime specification, which shortens a migration. Model names, voices, tools and some parameters still differ, so plan to rerun every test scenario before switching traffic. **Does the API handle phone calls on its own?** No. The API handles the voice. To answer or place a phone call you need a carrier in front of it. xAI's official example uses a Twilio phone number and Twilio Media Streams to pipe call audio to the API. **Which voices are available?** At launch, xAI announced three voices, Ara, Eve and Leo, plus inline tags such as \[whisper], \[sigh] and \[laugh] that let the agent sound more human. Check the current catalog before you build, because voices and models have changed several times since. **Where is the audio processed?** The real-time API documentation lists a single region, us-east-1, in the United States. If you serve callers in the EU or other regulated markets, raise data location, retention and training use with xAI before launch. **Is the API a good fit for outbound sales calls?** Be careful. In the US, the FCC ruled in February 2024 that AI-generated voices count as artificial voices under the TCPA, so the consent rules for artificial-voice calls apply. Most teams get more value from inbound use cases, such as reception, scheduling and FAQs. Keep reading ## Related articles [View all articles](https://skipcall.io/en/blog) 01 Voice AI ### [How to Connect Grok to a Real Phone Number](https://skipcall.io/en/blog/connect-grok-to-a-phone-number) A Grok voice agent is useless until it answers a real phone number. Here is how an AI reaches the phone network, the three ways to wire Grok to a number (an xAI number, a SIP trunk or Twilio Media Streams), how to keep the number your customers already know, and a go-live checklist. Written for people who do not write code. 3 Nov 2026 02 Voice AI ### [Grok Voice Agent Builder Review: A No-Code Phone Agent](https://skipcall.io/en/blog/grok-voice-agent-builder-review) Since July 1, 2026, xAI's Grok Voice Agent Builder turns a plain-English brief, a few documents and a calendar into an AI agent that answers a phone number. We reviewed what it really does, what a month of calls costs, which calls it handles well, and why it is still a beta you should pilot rather than roll out. 3 Nov 2026 03 Voice AI ### [How Grok Talks on the Phone: Voice AI, Explained](https://skipcall.io/en/blog/how-grok-talks-on-the-phone) Grok holds a phone conversation without the awkward pauses of older voice bots because one real-time model hears audio and speaks audio, with no detour through text. Here is how that works in plain English, why every AI voice still sounds like a phone call, which voices and languages Grok offers, and a five-minute test to judge any voice agent yourself. 3 Nov 2026 ## Ready to multiply your productivity? Join the sales teams that have transformed their prospecting. [Request a demo](https://skipcall.io/en/demo) Personalized demo • No commitment