# Bland Speech vs Grok TTS (2026): What a $25 per million price gap buys on a phone call

At $15 per million characters, Grok TTS costs less than half of what Bland Speech does, so a team has to decide what the other $25 per million is for. Grok TTS is the text to speech API from SpaceXSI (formerly xAI), and Design Arena's blind listening test of 4 August 2026 placed it 3rd of 13 AI voices, which makes it a serious option and not just a cheap one.

Bland Speech v3 scored 1365 in those same tests against 1156 for Grok TTS. The two also differ on what each was built around: SpaceXSI announced Grok TTS for voice agents, transcription, accessibility, podcasts, and interactive audio, and Bland built Bland Speech for phone calls.

## Which is better, Bland Speech or Grok TTS?

For a voice agent on a phone line where the caller should hear a person, Bland Speech is the better choice as of November 2026. Bland Speech v3 scored 1365 in Design Arena's Audio Realism Bench results of 4 August 2026, 209 points above Grok TTS in 3rd place at 1156, which is a wide gap on that scale. Grok TTS is the cheaper API by a large margin.

Grok TTS makes more sense for a team that wants the lower price per character, 20 languages from one API, or an invisible watermark on generated audio by default, and that already builds on SpaceXSI's other voice APIs. A team comparing more vendors can start with the [best text to speech for phone calls](/blog/best-tts-for-phone-calls).

<div class="cmp-usewhen"><div class="cmp-usewhen-own"><p><strong>Use Bland Speech when</strong></p><ul><li>Callers on a phone line should hear a voice that sounds like a person</li><li>A team wants pay as you go billing, with 50,000 free characters once for a new account</li><li>An existing OpenAI audio speech integration should switch providers by changing a base URL, a key, and a model name</li><li>A script needs performed delivery, such as a laugh or a warm read, written inline as tags</li><li>A voice agent runs on Pipecat or LiveKit Agents and needs a text to speech service built for those frameworks</li></ul></div><div><p><strong>Use Grok TTS when</strong></p><ul><li>A team wants a lower price per character and will pay $15 per million characters</li><li>A product speaks 20 languages from one API and wants character-level timestamps</li><li>A team needs an invisible AudioSeal watermark on generated audio, which SpaceXSI turns on by default</li><li>A team already builds on SpaceXSI's Voice APIs and wants speech, transcription, and a voice agent from one provider</li></ul></div></div>

| Feature | Bland Speech | Grok TTS |
| --- | --- | --- |
| Audio Realism Bench, 4 August 2026 | Scored 1365, ranked 1st of 13 AI voices | Scored 1156, ranked 3rd of 13 AI voices |
| Price per million characters | $40, one standard rate | $15, usage-based |
| First audio, as each company publishes it | 199ms | "Sub-second latency" for SpaceXSI's Voice APIs |
| Phone audio | Mu-law at 8 kHz, PCM from 8 to 48 kHz | Mu-law and A-law at 8 kHz, PCM, mp3, and wav |
| Delivery controls | Inline tags such as [laughs] and [say warmly], expressiveness and stability dials, timed pauses | Inline tags such as [laugh] and [pause], wrapping tags such as <whisper>, speed from 0.7 to 1.5 |
| Instant voice clone | One clean sample of about 10 seconds | A reference clip up to 120 seconds, with API access on Enterprise plans |

<details><summary>What these rows mean</summary>

- **Audio Realism Bench, 4 August 2026:** Audio Realism Bench is Design Arena's blind listening test, where native speakers hear two clips of the same line from two models and pick the one that sounds more like a real person. A higher score means listeners picked that voice more often. A recording of a real person scored 1500 in the same results.
- **Price per million characters:** Bland Speech costs $0.04 per 1,000 characters, which is $40 per million. SpaceXSI prices Text to Speech at $15.00 per million characters.
- **First audio:** Each is the company's own published claim. Bland publishes 199ms to first audio on a streamed request, and SpaceXSI describes its Voice APIs as a group as having "sub-second latency".
- **Phone audio:** US phone networks carry mu-law at 8 kHz, and both companies return it directly. SpaceXSI also returns A-law, the other G.711 phone format.
- **Delivery controls:** A tag is a word in brackets or angle brackets that tells the model how to speak the words around it.
- **Instant voice clone:** SpaceXSI says clips under 30 seconds may lack detail and recommends 90 to 120 seconds, and its console makes up to 30 custom voices for free.

</details>

## How do Bland Speech and Grok TTS compare?

The main difference is what each company set out to build. Bland Speech is a text to speech model made for phone calls, where, in Bland's words, polish reads as fake. Grok TTS is one part of SpaceXSI's voice stack, sold beside speech to text and a speech to speech agent API on the same account.

### Bland Speech overview

Bland Speech is the text to speech model inside Bland, a voice AI provider that builds and runs its own language, speech, and transcription models for enterprise phone calls. Bland Speech is sold on its own as an API and a studio on the [Bland Speech](/speech) page.

A team sends text over HTTP to POST /v2/tts and gets raw audio or a WAV file back, or streams language model tokens into a WebSocket at WSS /v2/tts/ws, where a new context interrupts the turn that is playing. POST /v2/audio/speech follows the OpenAI audio speech format, and a first-party Pipecat service and a LiveKit Agents plugin for Python are available. The [text to speech API](/text-to-speech-api) page has the code.

<div class="cmp-proscons cmp-proscons-own"><div class="cmp-proscons-pros"><p><strong>Pros</strong></p><ul><li>Bland Speech v3 ranked 1st of 13 AI voices on Audio Realism Bench on 4 August 2026, at 1365</li><li>Pay as you go at $0.04 per 1,000 characters, with no monthly plan</li><li>An OpenAI-compatible endpoint, so an existing integration moves over by changing a base URL</li><li>Performance tags, and a director in the studio for how a line is read</li><li>Mu-law at 8 kHz and PCM from 8 to 48 kHz, ready for phone calls</li><li>Instant clones from about 10 seconds of audio, and professional clones from 30 minutes or more</li></ul></div><div class="cmp-proscons-cons"><p><strong>Cons</strong></p><ul><li>More than a quick prototype needs, because Bland Speech is built for production calls</li></ul></div></div>

### Grok TTS overview

SpaceXSI (formerly xAI) sells Grok TTS as a standalone text to speech API inside its Grok API, announced on 17 April 2026. A request goes to POST /v1/tts, which returns audio bytes, or to a WebSocket where text arrives in pieces and audio comes back as base64 chunks. SpaceXSI says the same stack powers Grok Voice, Tesla vehicles, and Starlink customer support.

Grok TTS has 28 built-in voices, 20 languages through language codes, and a speed multiplier from 0.7 to 1.5. Pronunciation replacements take respellings or IPA, the phonetic alphabet linguists use. A setting called optimize_streaming_latency has three levels, and SpaceXSI says the lowest brings the fastest first audio with a more noticeable quality tradeoff at chunk boundaries.

<div class="cmp-proscons" data-rival="grok-tts"><div class="cmp-proscons-pros"><p><strong>Pros</strong></p><ul><li>$15 per million characters, less than half of Bland Speech's $40</li><li>Scored 1156 and ranked 3rd of 13 AI voices on Audio Realism Bench on 4 August 2026</li><li>20 languages through language codes, and more at varying accuracy</li><li>Mu-law and A-law output at 8 kHz for phone lines</li><li>An invisible AudioSeal watermark on generated audio by default</li><li>Character-level timestamps on request</li></ul></div><div class="cmp-proscons-cons"><p><strong>Cons</strong></p><ul><li>209 points behind Bland Speech v3 on Audio Realism Bench on 4 August 2026</li><li>Cloning through the API is gated to Enterprise plans</li><li>Custom voices are available in the United States only, excluding Illinois</li><li>The lowest-latency setting trades some quality at chunk boundaries</li></ul></div></div>

### How do Bland Speech and Grok TTS handle phone audio and latency?

Both return audio that a phone line carries without conversion, and they publish different kinds of latency claim. Bland publishes 199ms to first audio on a streamed request. SpaceXSI describes its Voice APIs as having "sub-second latency" as a group and gives Grok TTS its three-level streaming setting.

Bland Speech outputs mu-law at 8 kHz and PCM at 8, 16, 24, 44.1, or 48 kHz. Grok TTS outputs mp3, wav, pcm, mu-law, and A-law at 8, 16, 22.05, 24, 44.1, or 48 kHz. Both take text over a WebSocket as a language model writes it, and on Bland Speech a new context interrupts the turn that is playing.

### How do Bland Speech and Grok TTS handle security?

SpaceXSI says it is SOC 2 Type 2 compliant and sends a team that wants a business associate agreement (BAA), the contract HIPAA requires, to a questionnaire. SpaceXSI stores API requests and responses for 30 days, encrypted at rest, offers zero data retention, and says it never trains on API inputs or outputs without permission. SpaceXSI also says its US regional endpoint serves no voice APIs.

Bland is SOC 2 Type II audited, HIPAA compliant with a signed BAA, and FedRAMP 20x Class C certified at Moderate, granted on 17 November 2026. Speech synthesis runs inside that FedRAMP boundary, and Bland runs its own model and speech instead of wrapping a third party, as [Bland's trust and security record](/trust-security) sets out. Enterprise customers get dedicated orchestration and dedicated GPUs run by Bland.

## How much do Bland Speech and Grok TTS cost?

Grok TTS costs less than half as much per character, and both companies bill by usage. Bland Speech costs $40 per million characters, and Grok TTS costs $15 per million.

### Bland Speech pricing

Bland Speech costs $0.04 per 1,000 characters, directed or not, in the studio or the API. New accounts get 50,000 characters free, and credits do not expire. A 60-second note costs about $0.04, a 10-minute narration about $0.36, and a million characters $40.

Enterprise pricing is custom, set with Bland's sales team.

### Grok TTS pricing

SpaceXSI prices Text to Speech at $15.00 per million characters, with usage-based billing and, in SpaceXSI's words, "no hidden fees". Custom voices made in the SpaceXSI console are free up to 30 voices.

### What does a month of speech cost on each?

At 100,000 characters a month, about two hours of speech, Bland Speech costs $4 and Grok TTS costs $1.50. At 1 million characters, about 20 hours, the figures are $40 and $15. At 10 million characters, about 200 hours, they are $400 and $150.

Grok TTS costs less at every volume, and at 10 million characters a month the difference is $250, which a team weighs against a 209-point gap in blind listening tests.

## Which sounds more human, Bland Speech or Grok TTS?

In the blind tests Design Arena published on 4 August 2026, Bland Speech v3 scored 1365, 209 points above Grok TTS at 1156, which placed Grok TTS 3rd behind Microsoft's MAI-Voice-2 at 1224. Design Arena's own summary was that "Second only to human speech is Bland Speech v3 by Bland AI, with an Elo of 1365."

Elo is a rating built from head-to-head votes, so a higher score means listeners chose that voice as more human more often. Design Arena published the method, the prompt categories, and the voice choices in [an article on Audio Realism Bench](https://notes.designarena.ai/audio-realism-benchmark-measuring-realism-in-audio-models/). Each model was represented by one female and one male American English voice, so the scores describe those two voices and not all 28 of Grok TTS's, and most prompts ran up to about 40 words, with phone agent prompts in single turns.

## Why is Bland Speech the best Grok TTS alternative?

A team that picked Grok TTS for its price and now weighs how a caller hears it is the buyer Bland Speech fits, because Bland Speech v3 scored 1365 against 1156 in the blind tests of 4 August 2026 and returns mu-law at 8 kHz straight from the API.

A team on OpenAI's audio speech API can move by changing a base URL, and a team that wants a brand voice can clone one from about 10 seconds of audio, where SpaceXSI recommends 90 to 120 seconds for its best results. A professional clone trains on 30 minutes or more of audio after Bland's consent and rights review, and every clone stays private to the organization that made it.

Bland's [comparison of text to speech APIs](/compare/text-to-speech) covers more vendors, and the [head-to-head comparisons](/compare/text-to-speech/head-to-heads) collect the pairs. A team that cares about speed can read [time to first audio](/blog/tts-latency-explained), and one building an agent can start with the [best text to speech APIs for voice agents](/blog/best-tts-for-voice-agents).

## Frequently asked questions

### Is Bland Speech better than Grok TTS?

For phone calls and voice agents where the voice should sound like a person, Bland Speech is better than Grok TTS. Bland Speech v3 scored 1365 on Audio Realism Bench on 4 August 2026, and Grok TTS ranked 3rd at 1156. Grok TTS costs $15 per million characters against Bland's $40.

### Is Grok TTS cheaper than Bland Speech?

Yes. SpaceXSI charges $15.00 per million characters for Grok TTS, and Bland Speech costs $40 per million, so a month of 10 million characters comes to $150 on Grok TTS and $400 on Bland Speech.

### Who makes Grok TTS?

SpaceXSI makes Grok TTS. SpaceXSI was formerly xAI and still calls itself "SpaceXSI (xAI)" to developers. Grok TTS launched as a standalone text to speech API in 2026.

### Does Grok TTS clone voices?

Yes. SpaceXSI clones a voice from a reference clip up to 120 seconds long, and recommends 90 to 120 seconds for the best results. The console makes up to 30 custom voices free, API access is gated to Enterprise plans, and custom voices are available in the United States only, excluding Illinois.

### Which is faster, Bland Speech or Grok TTS?

Each company publishes a different kind of claim. Bland publishes 199ms to first audio on a streamed request, and SpaceXSI describes its Voice APIs as having sub-second latency, with each figure coming from its own company's method.

### Does Grok TTS work on phone calls?

Yes. Grok TTS returns mu-law and A-law at 8 kHz, the G.711 formats that phone networks carry.

### Does Bland Speech work on phone calls?

Yes. Bland built Bland Speech for phone calls, and it returns mu-law at 8 kHz as well as PCM from 8 to 48 kHz, so audio goes onto a phone line without conversion.
