A store that answers 100,000 support calls a month and reads back about 1,500 characters on each pays for text to speech by the character, and one of those replies costs anywhere from six tenths of a cent to fifteen cents depending on which API reads it. That gap is wide enough that a store should decide what a caller hears for the extra money before it picks a vendor.
Bland Speech takes first place by pairing one standard rate of $0.04 per 1,000 characters, which is six cents for that reply, with a voice that ranked first of 13 AI voices in Design Arena's Audio Realism Bench results of 4 August 2026. Cheaper options exist, since Amazon Polly's Standard voices cost a tenth as much, and every price here is each company's own as of November 2026.
What are the best text to speech APIs for e-commerce support calls?
The best text to speech API for e-commerce support calls is Bland Speech, for stores that want a realistic voice at one standard rate: Bland Speech v3 ranked first of 13 AI voices on Audio Realism Bench on 4 August 2026, and a 1,500-character reply costs six cents at $0.04 per 1,000 characters.
- Bland Speech: Best overall for store support lines that want a realistic voice at one standard rate
- ElevenLabs: Best for stores that want self-serve voice cloning and 90+ languages
- Rime: Best for stores that want billing by the minute and a spell() function for order codes
- Inworld: Best for stores that want a low rate and delivery steered in plain words
- Murf: Best for stores that want a low flat rate on streaming speech
- Smallest AI: Best for stores that want a low rate with mu-law output
- Amazon Polly: Best for stores that replay the same lines all day at $4 per million characters
What these rows mean
- Price per million characters: list prices as each company publishes them, with the model or engine range shown. Rime bills by the minute instead.
- Cost of a 1,500-character reply: list price multiplied out, with Rime's minute counted as about 1,000 characters, as Rime says. Plan allowances and volume discounts aren't included.
- Audio Realism Bench: the rank among 13 AI voices and the score in Design Arena's published results.
At 100,000 replies a month the speech bill runs from $600 to $15,000 across these list prices, so the real choice is what a caller hears when an order has gone wrong, which is what Design Arena's published results of 4 August 2026 measured in blind listening.
What should an online store check in a text to speech API?
A store should check how the API reads order numbers and product names, whether it can speak in the store's own voice, and whether audio arrives in a format phone lines carry.
How does the API read order numbers and product names?
A support call is mostly strings read back: an order number, a tracking code, a size, a return window. Rime has a spell() function for letter-by-letter reading, Smallest AI takes pronunciation dictionaries for brand names and acronyms, Inworld and Murf take inline IPA and pronunciation dictionaries, and Amazon Polly takes SSML phonetic tags. Bland Speech documents timed pauses written as <|0.3|>, which can separate groups of digits, plus performance tags for delivery.
Can the store use its own brand voice?
Bland Speech makes an instant clone from one clean sample of about 10 seconds, and a professional clone is fine-tuned on 30 minutes or more of audio and goes through Bland's consent and rights review. Inworld clones from as little as 3 seconds on every plan, ElevenLabs offers instant clones from its $6 Starter plan, and Smallest AI clones from 5 to 15 seconds. Rime and Murf reserve cloning for enterprise customers, and Amazon Polly builds a Brand Voice as an engagement through an AWS account manager.
Does the audio fit a phone line?
Mu-law at 8 kHz is the format North American phone networks carry. Bland Speech, ElevenLabs, Rime, Inworld, Smallest AI, and Amazon Polly document mu-law output, and Murf names A-law and mu-law among its HTTP streaming formats.
How do the 7 best text to speech APIs for e-commerce support calls compare?
These seven differ most in price per reply and in how each voice sounds to a caller.
1. Bland Speech: Best overall for store support lines that want a realistic voice at one standard rate
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 for enterprise phone calls. Text streams in over HTTP at POST /v2/tts or over a WebSocket that takes text as a language model writes it, and audio comes back as PCM from 8 to 48 kHz or as mu-law at 8 kHz. Bland's text to speech API page has the endpoints, and Bland Speech suits stores putting a voice on order status, returns, and delivery calls.
Pros
- Bland Speech v3 ranked first of 13 AI voices on Audio Realism Bench on 4 August 2026
- One standard rate, $0.04 per 1,000 characters, with no monthly plan
- Timed pauses and performance tags such as [say warmly] shape each read
- Instant clones from about 10 seconds of audio, private to the customer's organization
Cons
- More than an FAQ widget needs, because Bland Speech is built for production phone calls
Pricing: $0.04 per 1,000 characters, which is $40 per million, pay as you go with no monthly plan, and 50,000 characters free once for a new account. Enterprise pricing is custom, set with Bland's sales team. The Bland Speech page has a studio to try a voice.
2. ElevenLabs: Best for stores that want self-serve voice cloning and 90+ languages
What does it do? ElevenLabs sells text to speech, transcription, and voice agents. Eleven v4 and v4 Turbo launched on 28 September 2026, the v4 family covers 90+ languages, and audio tags such as [whispering] and [laughing] shape delivery on Eleven v4. The API returns mu-law at 8 kHz.
Pros
- 90+ languages on the Eleven v4 family
- Instant clones from the Starter plan at $6 a month
- v4 Turbo and Flash at $0.04 per 1,000 characters
Cons
- Eleven v3 ranked 12th of 13 AI voices on Audio Realism Bench on 4 August 2026
- Eleven v4 costs $0.08 per 1,000 characters
- Free plan use is non-commercial, with a commercial license from Starter
Pricing: $0.08 per 1,000 characters for Eleven v4, v3, and Multilingual v2, and $0.04 for v4 Turbo, v3 Conversational, and Flash. Eleven v4 and v4 Turbo came out after those results and aren't scored in them.
3. Rime: Best for stores that want billing by the minute and a spell() function for order codes
What does it do? Rime sells conversational voice models, with Coda and Mist v3 for text to speech, and says they are built for voice agents, IVR, and telephony. They return G.711 mu-law at 8 kHz.
Pros
- A spell() function for letter-by-letter reading of codes
- $50 of free usage every month on the Free plan
- SOC 2 Type 2 and HIPAA compliance, with a BAA available
Cons
- Coda's cloud API serves 10 languages
- Voice cloning is an enterprise order placed with Rime
- Its voices weren't in Audio Realism Bench's results of 4 August 2026
Pricing: Pay as you go at $0.030 a minute for Mist v3 and $0.050 a minute for Coda v2, and Rime says a minute is about 1,000 characters.
4. Inworld: Best for stores that want a low rate and delivery steered in plain words
What does it do? Inworld sells realtime voice AI aimed at consumer-facing applications. Realtime TTS-2 launched on 5 May 2026 and takes delivery instructions in plain words, Realtime TTS-2 Flash launched on 9 August 2026 as its fastest and cheapest model, and a speech endpoint in OpenAI's format shipped on 11 September 2026.
Pros
- TTS-2 Flash at $15 per million characters on pay as you go, falling to $7 on the Growth plan
- Instant clones from as little as 3 seconds of audio on every plan
- SOC 2 Type II, with a commercial license on the $0-a-month On-Demand plan
Cons
- TTS-1.5 Max ranked 13th of 13 AI voices on Audio Realism Bench on 4 August 2026, and Realtime TTS-2 wasn't in those results
- Plain-words steering works on TTS-2 only, not on Flash
Pricing: $25 per million characters for TTS-2 and $15 for TTS-2 Flash on pay as you go, down to $12.50 and $7 on the $1,500-a-month Growth plan.
5. Murf: Best for stores that want a low flat rate on streaming speech
What does it do? Murf sells a text to speech API with two models: Falcon 2.1, launched on 19 October 2026 for streaming, and Gen2 for studio quality. Murf says Falcon's time to first audio is under 100ms, and it offers 150+ voices in 35 languages.
Pros
- Falcon 2.1 at $0.01 per 1,000 characters
- Purchased credit never expires, with $10 of free API credit to start
- SOC 2 Type II and ISO 27001, with reports under NDA
Cons
- Falcon 2.1 voices speak in a single style, Conversational
- Instant voice cloning is Enterprise only
- Its voices weren't in Audio Realism Bench's results of 4 August 2026
Pricing: $0.01 per 1,000 characters for Falcon 2.1, which is $10 per million, on pay as you go.
6. Smallest AI: Best for stores that want a low rate with mu-law output
What does it do? Smallest AI sells a realtime voice suite for enterprise, and Lightning v3.1 Pro is its current text to speech model. Smallest AI says first audio arrives in under 100ms, and the API returns PCM, WAV, MP3, or mu-law from 8 kHz up.
Pros
- About $19.50 per million characters for Lightning v3.1 Pro
- $10 in free credits on pay as you go
- Pronunciation dictionaries for brand names and acronyms
Cons
- Lightning v3.1 Pro ranked 11th of 13 AI voices on Audio Realism Bench on 4 August 2026
- Lightning v3.1 retires on 30 November 2026
- HIPAA zero data retention is a $1,000-a-month add-on on pay as you go
Pricing: About $17.50 per million characters for Lightning v3.1 and $19.50 for v3.1 Pro on pay as you go.
7. Amazon Polly: Best for stores that replay the same lines all day at $4 per million characters
What does it do? Amazon Polly offers four voice engines, Generative, Long-Form, Neural, and Standard, streams audio over HTTP, and returns mu-law and A-law at 8 kHz.
Pros
- Standard voices at $4 per million characters
- Cached audio replays at no additional cost
- SSML controls for pauses, pace, emphasis, and phonetic pronunciation
Cons
- Generative voices can shift slightly between model updates, as Amazon warns
- No self-serve voice cloning, with Brand Voice as an engagement through an AWS account manager
- Its voices weren't in Audio Realism Bench's results of 4 August 2026
Pricing: $4 per million characters for Standard voices, $16 for Neural, $30 for Generative, and $100 for Long-Form, with free tier allowances by engine.
How much does text to speech cost on a store support line?
Text to speech APIs charge per character or per minute, and one reply of about 1,500 characters, roughly a minute and a half of speech, costs between $0.006 and $0.15 at list price. At 100,000 such replies a month, 150 million characters, the monthly speech bill comes to:
- Bland Speech at $40 per million: $6,000
- Amazon Polly Standard at $4 per million: $600, and Neural at $16: $2,400
- Murf Falcon 2.1 at $10 per million: $1,500
- Inworld Realtime TTS-2 Flash at $15 per million: $2,250, and TTS-2 at $25: $3,750
- Smallest AI Lightning v3.1 Pro at about $19.50 per million: $2,925
- Rime Mist v3 at $0.030 a minute: $4,500, and Coda v2 at $0.050: $7,500
- ElevenLabs Eleven v4 Turbo at $40 per million: $6,000, and Eleven v4 at $80: $12,000
Bland Speech sits mid-table. Polly Standard costs a tenth as much and Murf's Falcon a quarter, and neither was in Audio Realism Bench's results of 4 August 2026, so a $5,400 monthly gap against Polly Standard buys a voice that ranked first of 13 in blind listening that day. Those are list prices before plan allowances and volume discounts.
Lines that never change, such as a shipping confirmation, are where Polly's free replay of cached audio helps most.
How were these text to speech APIs compared?
Samuel Oram-Jones compared seven APIs on price per reply, how each reads order numbers and product names, brand voice options, phone audio, and voice realism. Rival details come from each company's own pages, and voice scores from Design Arena's results of 4 August 2026.
Bland's comparison of text to speech APIs covers the same vendors by job, and the best lists group them by use. A store weighing price first can read text to speech API pricing, compared, and Bland Speech vs Rime puts Bland Speech beside Rime.
Frequently asked questions
What is the best text to speech API for e-commerce?
Bland Speech is the best text to speech API for e-commerce support calls in 2026 for stores that put voice quality first, at $0.04 per 1,000 characters on its standard rate.
How much does text to speech cost per support call?
A reply of about 1,500 characters costs $0.06 on Bland Speech, between $0.06 and $0.12 on ElevenLabs, $0.015 on Murf's Falcon 2.1, and $0.006 on Amazon Polly Standard voices, at list prices.
What is the cheapest text to speech API for a store?
Amazon Polly Standard voices cost $4 per million characters and Murf's Falcon 2.1 costs $10 per million, the lowest list prices here. Neither was in Audio Realism Bench's results of 4 August 2026, and Bland Speech costs $40 per million.
Can a text to speech API read order numbers correctly?
Each API gives a store different tools, from Rime's spell() function to Bland Speech's timed pauses, and a store should test its own order and tracking numbers on each before choosing.
Can a store use a cloned brand voice?
Yes. Bland Speech, Inworld, ElevenLabs, and Smallest AI clone from a short sample on self-serve terms, while Rime, Murf, and Amazon Polly keep custom voices for enterprise or account-managed work.
