# 6 best text to speech APIs without a monthly subscription in 2026

By Samuel Oram-Jones

A monthly text to speech plan is a bet on next month's volume, and the vendor keeps the bets that lose. Gradium says credits from its plans must be used within the current month and cannot roll over, and Cartesia lets credits roll over only up to 2x a plan's monthly allowance, so a team that records a batch of prompts in March and nothing in April pays for April anyway.

Six text to speech APIs skip the plan: Bland Speech, Deepgram, Murf, Fish Audio, Inworld, and Rime. Bland Speech comes first here because Bland Speech v3 ranked first among AI voices in a blind listening test of 4 August 2026, it returns phone-ready audio, and its credits never expire. Every rate and plan here is as each company publishes it as of November 2026.

## Which text to speech APIs have no monthly subscription?

Bland Speech, Deepgram, Murf, Fish Audio, Inworld, and Rime each sell text to speech with no monthly fee on the tier named in each row, and Bland Speech is the one to pick when a voice has to sound like a person on a phone call.

- **Bland Speech:** Best overall for phone calls and voice agents, on a balance that never expires
- **Murf:** Best for the lowest published rate here, with purchased credit that never expires
- **Fish Audio:** Best for developers who want a free model to build on and $15 per million UTF-8 bytes in production
- **Deepgram:** Best for teams that want $200 of free credit with no expiry and speech recognition from the same vendor
- **Inworld:** Best for consumer apps that want steerable voices and a commercial license at $0 a month
- **Rime:** Best for voice agents that pay by the minute, with $50 of free usage every month

| API | Monthly fee | Published rate | One million characters | Audio Realism Bench, 4 August 2026 |
|---|---|---|---|---|
| Bland Speech | None | $0.04 per 1,000 characters | $40 | #1 AI voice, 1365 |
| Deepgram | None on pay as you go | Flux TTS $0.045 per 1,000 characters | $45 | Wasn't in those results |
| Murf | None on API pay as you go | Falcon 2.1 $0.01 per 1,000 characters | $10 | Wasn't in those results |
| Fish Audio | None for API access | s2.1-pro $15 per million UTF-8 bytes | About $15 for English text | Wasn't in those results |
| Inworld | None on On-Demand | Realtime TTS-2 $25 per million characters | $25 | TTS-2 wasn't in those results |
| Rime | None on Flex | Coda v2 $0.050 a minute | About $50 | Wasn't in those results |

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

- **Monthly fee:** Whether the API sells access with no plan, on the tier named in the cell.
- **Published rate:** The rate each company publishes for its main model, in the unit it uses. Fish Audio counts UTF-8 bytes and Rime counts minutes.
- **One million characters:** The published rate multiplied out. A UTF-8 byte is one character in English text, and Rime counts about 1,000 characters a minute.
- **Audio Realism Bench, 4 August 2026:** Design Arena's blind listening test, where listeners hear two recordings of the same line and pick the one that sounds more like a real person. A human recording scored 1500.

</details>

Two of these APIs cost less than Bland Speech per million characters, since Murf publishes $10 and Fish Audio $15 per million bytes, so first place rests on the voice, which ranked first of 13 AI voices in [Design Arena's Audio Realism Bench results](https://notes.designarena.ai/audio-realism-benchmark-measuring-realism-in-audio-models/) of 4 August 2026, and on audio a phone line can use as it arrives.

## What does a one-off batch of speech cost without a plan?

A one-off job of 2 million characters costs $20 on Murf, about $30 on Fish Audio, $50 on Inworld, $80 on Bland Speech, $90 on Deepgram, and about $100 on Rime, with nothing owed before or after. Idle months cost nothing, and unspent credit stays put: Bland says its credits do not expire, Murf says purchased credit never expires, and Deepgram says its pay as you go plan has no expiration.

### How does that compare with a plan?

Cartesia's Startup plan is $49 a month for 1.25 million credits, and a credit pays for one character. The 750,000 characters past the allowance bill at $45 per million if a team switches overages on, so the same job costs about $83 in its month and another $49 for each further month the plan stays active. Bland Speech's $80 lands close to that, so the saving shows up in the idle months, and Murf's $20 is a quarter of Bland Speech's price for a team whose only goal is the smallest bill.

## How do the 6 best text to speech APIs without a subscription compare?

The six skip the plan in different ways: a wallet at Bland Speech, a credit at Deepgram and Murf, a bytes meter at Fish Audio, an On-Demand tier at Inworld, and a per-minute Flex rate at Rime.

### 1. Bland Speech: Best overall for phone calls and voice agents, on a balance that never expires

**What does it do?** Bland Speech is the text to speech model inside Bland, a voice AI provider that runs its own model and speech, and it was built for phone calls. It streams over HTTP at POST /v2/tts or over a WebSocket at /v2/tts/ws, which takes text as a language model writes it, and it returns PCM from 8 to 48 kHz or mu-law at 8 kHz.

**Who is it for?** Bland Speech suits teams putting a voice on phone calls and voice agents, and teams whose volume arrives in bursts, because a loaded balance doesn't lapse.

<div class="cmp-proscons cmp-proscons-own"><div class="cmp-proscons-pros"><p><strong>Pros</strong></p><ul><li>Bland Speech v3 ranked first of 13 AI voices on Audio Realism Bench on 4 August 2026</li><li>One standard rate with no monthly plan, and credits that do not expire</li><li>50,000 characters free for a new account</li><li>Mu-law and PCM from 8 to 48 kHz, ready for phone calls</li></ul></div><div class="cmp-proscons-cons"><p><strong>Cons</strong></p><ul><li>More than a hobby project needs, because Bland Speech is built for production calls</li></ul></div></div>

**Pricing:** $0.04 per 1,000 characters, which is $40 per million, in the studio or the API, billed against a wallet balance, with 50,000 characters free to a new account. Each request carries a minimum charge of $0.001, and some public-library voices add a per-character creator fee. Enterprise pricing is custom, set with Bland's sales team. The [Bland Speech](/speech) page has a studio to try it.

### 2. Deepgram: Best for teams that want $200 of free credit with no expiry and speech recognition from the same vendor

**What does it do?** Deepgram sells speech recognition and text to speech. Flux TTS, its current model, streams over a WebSocket at /v2/speak, and Deepgram says responses start in as low as 80ms.

**Who is it for?** Deepgram suits teams already transcribing calls with Deepgram that want speech from the same account.

<div class="cmp-proscons" data-rival="deepgram"><div class="cmp-proscons-pros"><p><strong>Pros</strong></p><ul><li>$200 of free credit with no card and no expiry</li><li>Pay as you go has no minimums</li><li>Mu-law and A-law at 8 kHz</li></ul></div><div class="cmp-proscons-cons"><p><strong>Cons</strong></p><ul><li>The Growth plan, with lower rates, is $4,000 a year prepaid</li><li>Deepgram's voices weren't in Audio Realism Bench's results of 4 August 2026</li></ul></div></div>

**Pricing:** $0.045 per 1,000 characters for Flux TTS, $0.030 for Aura-2, and $0.015 for Aura-1 on pay as you go. Growth starts at $4,000 a year in prepaid credits and prices Aura-2 at $0.027, and Flux TTS earns $1 in credits for every $1 spent, up to $500, through 31 December 2026.

### 3. Murf: Best for the lowest published rate here, with purchased credit that never expires

**What does it do?** Murf's API sells two text to speech models, Falcon 2.1 for real-time agents and Gen2 for studio-style narration. Murf says Falcon 2.1 delivers first audio in under 100ms.

**Who is it for?** Murf suits teams generating large volumes of speech, such as batch narration and prompt libraries, that want credit with no expiry date.

<div class="cmp-proscons" data-rival="murf"><div class="cmp-proscons-pros"><p><strong>Pros</strong></p><ul><li>Falcon 2.1 at $0.01 per 1,000 characters, the lowest published rate here</li><li>Purchased credit never expires</li><li>$10 of free API credit with no card</li></ul></div><div class="cmp-proscons-cons"><p><strong>Cons</strong></p><ul><li>Falcon 2.1 voices speak in a single style, Conversational</li><li>Instant voice cloning is on the Enterprise plan only</li><li>Murf's voices weren't in Audio Realism Bench's results of 4 August 2026</li></ul></div></div>

**Pricing:** Falcon 2.1 is $0.01 per 1,000 characters ($10 per million) and Gen2 is $0.03 per 1,000 characters on API pay as you go. [Bland Speech vs Murf](/compare/bland-speech-vs-murf) compares the two side by side.

### 4. Fish Audio: Best for developers who want a free model to build on and $15 per million UTF-8 bytes in production

**What does it do?** Fish Audio sells text to speech through a web app and a REST and WebSocket API. S2.1-Pro is its recommended production model, and s2.1-pro-free is the same model at $0 for development under fair-use limits.

**Who is it for?** Fish Audio suits developers who want to build against a free model and pay for production traffic by the byte.

<div class="cmp-proscons" data-rival="fish-audio"><div class="cmp-proscons-pros"><p><strong>Pros</strong></p><ul><li>No subscription fees or monthly minimums for API access</li><li>A free model for development</li><li>s2.1-pro supports 83 languages</li></ul></div><div class="cmp-proscons-cons"><p><strong>Cons</strong></p><ul><li>Billing counts UTF-8 bytes rather than characters</li><li>Output formats are MP3, WAV, PCM, and Opus</li><li>Professional voice clones are made in the web app on a paid app plan</li><li>Fish Audio's voices weren't in Audio Realism Bench's results of 4 August 2026</li></ul></div></div>

**Pricing:** $15.00 per million UTF-8 bytes for s2.1-pro, which Fish Audio converts to about 180,000 English words or 12 hours of speech.

### 5. Inworld: Best for consumer apps that want steerable voices and a commercial license at $0 a month

**What does it do?** Inworld sells realtime voice AI for consumer apps, with Realtime TTS-2 and Realtime TTS-2 Flash. Both stream over HTTP and a WebSocket that detects sentence boundaries in text from a language model, return mu-law and A-law for phone lines, and take 50 non-verbal sounds such as [laugh] and [sigh].

**Who is it for?** Inworld suits teams building companions, games, and consumer voice agents that want steerable delivery and a commercial license without a plan.

<div class="cmp-proscons" data-rival="inworld"><div class="cmp-proscons-pros"><p><strong>Pros</strong></p><ul><li>On-Demand costs $0 a month and includes a commercial license and up to 70 minutes of TTS</li><li>Instant cloning from as little as 3 seconds of audio on every plan</li><li>A speech endpoint in OpenAI's format</li></ul></div><div class="cmp-proscons-cons"><p><strong>Cons</strong></p><ul><li>Steering in plain words works on TTS-2 only</li><li>Lower rates sit on monthly tiers that start at $25</li><li>Realtime TTS-2 wasn't in Audio Realism Bench's results of 4 August 2026</li></ul></div></div>

**Pricing:** On-Demand is $0 a month with TTS-2 at $25 per million characters and TTS-2 Flash at $15 per million.

### 6. Rime: Best for voice agents that pay by the minute, with $50 of free usage every month

**What does it do?** Rime sells conversational voice models: Coda, its flagship text to speech model, and Mist v3, a low-latency one. Text to speech streams over HTTP, WebSockets, and gRPC. Output includes G.711 mu-law at 8 kHz.

**Who is it for?** Rime suits teams building voice agents that prefer to budget by the minute of speech rather than by the character.

<div class="cmp-proscons" data-rival="rime"><div class="cmp-proscons-pros"><p><strong>Pros</strong></p><ul><li>Flex has no monthly fee</li><li>$50 of free usage every month, renewing monthly</li></ul></div><div class="cmp-proscons-cons"><p><strong>Cons</strong></p><ul><li>Coda v2 costs $0.050 a minute, about $50 per million characters on Rime's own conversion</li><li>Voice cloning is an enterprise order that needs 30 to 60 minutes of recordings</li><li>Rime's voices weren't in Audio Realism Bench's results of 4 August 2026</li></ul></div></div>

**Pricing:** Flex is billed by the minute, with Mist v3 at $0.030, Coda v2 at $0.050, and Prism at $0.080, and Rime counts about 1,000 characters to a minute.

## Which text to speech APIs still charge a monthly subscription?

Cartesia, ElevenLabs, and Gradium sell text to speech through monthly plans.

Cartesia's Pro plan is $5 a month for 100,000 credits, Startup is $49 for 1.25 million, and Scale is $299 for 8 million, while its Free plan gives 20,000 credits a month for evaluation and personal use. ElevenLabs runs plans from Starter at $6 a month to Business at $990, with 10,000 v4 characters a month on Free. Gradium's plans run from XS at $12 a month on its default billing for 225,000 credits to L at $1,480.

A plan is not always the worse deal. Cartesia's Scale plan works out to $37.38 per million credits, below Bland Speech's $40, and Gradium prices its L plan at $36 per million credits. A plan pays off for a team that speaks a similar amount every month, and the six APIs here pay off for a team whose months look different from each other. [Text to speech API pricing in 2026, compared](/blog/text-to-speech-api-pricing) sets the rates side by side.

## How were these text to speech APIs compared?

The six were compared on a tier with no monthly fee, each voice's result in Design Arena's Audio Realism Bench of 4 August 2026, the published rate multiplied to a million characters, what unspent credit does, and whether the audio comes out in a format a phone line takes.

Companies that no longer sell text to speech were left out. The [text to speech comparison hub](/compare/text-to-speech) lists every provider, the [best lists](/compare/text-to-speech/best-lists) page ranks them on other questions, and the [text to speech API](/text-to-speech-api) page has request examples.

## Frequently asked questions

### Which text to speech API has no monthly subscription and sounds the most human?

Bland Speech: it bills $0.04 per 1,000 characters pay as you go with no monthly plan, and Bland Speech v3 ranked first of 13 AI voices in Design Arena's Audio Realism Bench results of 4 August 2026.

### Does Bland Speech charge a monthly fee?

No. Bland Speech bills each character against a wallet balance at $0.04 per 1,000 characters, and a new account gets 50,000 characters free. Credits do not expire, and Bland's $0.001 minimum per request and some library voices' creator fees still apply.

### What is the cheapest text to speech API with no subscription?

Murf, at $0.01 per 1,000 characters for Falcon 2.1, which is $10 per million characters. Fish Audio follows at $15 per million UTF-8 bytes, which is about $15 per million characters in English text. Bland Speech costs $40 per million and is chosen for the voice, not the price.

### Does a text to speech free tier count as no subscription?

A free tier is no subscription only while a team stays inside it. ElevenLabs' Free plan is non-commercial and Cartesia's Free plan is for evaluation and personal use, so a team shipping to customers moves to a paid monthly plan at each. Deepgram's $200 credit and Murf's $10 credit are balances rather than plans, so they need no monthly fee at all.

### Can a text to speech API with no subscription handle phone calls?

Yes. Bland Speech returns mu-law at 8 kHz or PCM from 8 to 48 kHz, Deepgram returns mu-law and A-law at 8 kHz, and Rime returns G.711 mu-law at 8 kHz. Murf's HTTP streaming returns A-law and mu-law.

### Can an OpenAI text to speech integration move to Bland Speech?

Yes. A team changes a base URL, an API key, and a model name, and the [migration guide from OpenAI TTS](/blog/migrate-from-openai-tts) walks through the steps.
