Introducing Simba 3.5
Spoken like a local in 48 languages. Expressive text-to-speech and voice cloning, through one API.
Looking for the Speechify reading app? It lives at speechify.com
#1
Simba 3.5 among 51 models, read 13 Nov 2026. On this board every model speaks in the same eight cloned voices, four US and four UK English, judged in blind votes. Its 95% interval overlaps the models closest to it, so the rank is a dated reading, not a guaranteed quality advantage.
View the board72ms
Simba 3.5's time to first audio for 95% of streaming requests, under production load
The Simba 3.5 fact sheet's English results against Simba 3.2, as of October 2026. A 95th percentile under production load, not a median or a guarantee.
Read the fact sheetList price $15 launch price $10/ 1M
Per million characters of Simba 3.5 on Starter, launch price until 21 Nov
Double usage on the fee, in one shared monthly balance, not a separate character allowance: it buys 2M characters of Simba 3.5 until 21 Nov 2026, 1.3M after1.3M characters of Simba 3.5, or 2M characters of Simba 3.0 and 3.2. Continued usage draws from your top-up balance.
Compare plansIndependent blind preference
Artificial Analysis Controlled Voice
Our models on the same eight voices: Elo and 95% confidence interval
On this board every model speaks in the same eight cloned voices, four US and four UK English, judged in blind votes, so the voice catalogue is out of the comparison. Bars show each model's 95% confidence interval. Simba 3.5's overlaps the models closest to it on the board, so its #1 is a dated reading, not a guaranteed quality advantage.
View the boardArtificial Analysis runs blinded pairwise listening tests. On its Controlled Voice board, read 13 Nov 2026, Simba 3.5 scored 1,192 Elo with a 95% interval of 1,175 to 1,209, #1 of 51 models with a rank range of 1 to 3.
On its full provider-voice board, read 13 Oct 2026, Simba 3.2 scored 1,244 ±13 among 96 models, with the lowest Artificial Analysis price in the top ten. The dated figures are SpeechifyAI's reviewed transcriptions, not official AA exports.
Production US East
Streaming response
Time to first audio byte on our production path
Measured first-byte latency, not audible latency, end-to-end agent response time, or a cross-vendor benchmark.
These figures measure our Simba 3.2 production streaming path in US East from request to the first audio byte. Network distance, input length, client buffering, and the boundary a benchmark chooses can all change the result.
Time to first byte and time to first audible sound are different measurements. We publish this as a dated production observation, not a guarantee or an independent claim that every route is fastest.
The board's top ten
The board's top ten, compared on one cost model
USD per 1M characters, zero-based scale
AA-normalized estimates across the top ten of the 13 Oct 2026 provider-voice board. Subscription assumptions apply. Not a quote or total monthly bill.
Inspect the dated figuresArtificial Analysis normalizes API pricing to an estimated per-character measure. For subscriptions, it uses the lowest-cost annual plan with at least 1M characters and assumes 80% usage. This chart compares only the board's top ten by Elo.
The $6.60 figure is that independent model, not a SpeechifyAI billed rate. Our current subscription fees, included usage, and subsequent usage rates are published separately.
Voice cloning. 48 languages.
Explore what makes a voice feel like a voice.
Simba 3.5 · Voice cloning
Start with a consented reference recording. Create a reusable voice for the words that come next.
Explore voice cloning01 / Reference
02 / Clone
Two different recordings. Generated sample: Simba 3.5. Only clone a voice you have permission to use.
Simba 3.5 · Multilingual synthesis
Hear one line in 6 of Simba 3.5's 48 languages, each read by a voice native to it.
Explore language supporten-GB · Imogen
de-DE · Henrik
es-ES · Lola
fr-FR · Augustin
it-IT · Vittoria
pt-BR · Letícia
Simba 3.5Set language on the request and any catalog voice speaks it, on one streaming-native model id.
Flagship · 48 languages · Voice cloning · 24kHz audio
Simba 3.2An older English model that remains available. Set pauses and pacing with SSML, and keep the character of the voice.
Older model · English · Voice cloning
Simba 3.0An older multilingual model that remains available for English, German, Spanish, French, Italian, and Portuguese.
Older model · 6 languages · 7 locales · 24kHz audio
Send text. Receive speech. One streaming API, with control over voice and timing.
curl https://api.speechify.ai/v1/audio/stream \
-H "Authorization: Bearer $SPEECHIFY_API_KEY" \
-H "Content-Type: application/json" \
-H "Accept: audio/mpeg" \
-d '{
"input": "Your next idea. Now with a voice.",
"model": "simba-3.5",
"language": "en-US",
"voice_id": "harper_32"
}' --output speech.mp3Build and ship on the text-to-speech API. No credit card.
$0/month
Start freeFor solo devs and early-stage projects.
$10/month
Start BuildingFor teams shipping speech in production.
$99/month
Start ProFor serious production speech traffic.
$499/month
Start ScaleBeyond the standard plans
For global volume, compliance, and procurement.
Simba 3.5 lists at $15/$12/$10 per 1M characters on Starter/Pro/Scale, billed at $10/$8/$6 until 21 Nov 2026. Simba 3.0 and 3.2 bill $10/$8/$6 per 1M characters. No card needed to start. Voice cloning on paid plans. Included usage renews monthly; purchased top-ups never expire.
Product updatesUpdated Nov 13, 2026
48languages on one model, Simba 3.5
Product updatesUpdated Nov 13, 2026
Build guidesOct 15, 2026
Limit one end user, not your whole workspace.
Build guidesOct 15, 2026
Quality, price and speed, with dated numbers.
How good does it sound?
1,244Elo
What does it cost?
$91a month for 10M on Starter
How fast does it start?
47ms to first audible audio over the WebSocket
Use Simba 3.5, the recommended model: one model id speaks 48 languages. Simba 3.2 and 3.0 stay available. Check each voice's supported models before choosing a pairing.
Explore the model familyYes. Voice cloning on paid plans creates a reusable voice from a reference recording. Use a voice you own or have the speaker's permission to clone, and check the voice's supported models before synthesis.
Explore voice cloningThe listening demos use prepared recordings. In the SpeechifyAI platform you can try your own text, explore voices, and get an API key for your application.
See how the API worksText-to-speech is metered by character, at the rate of the model you call. Starter is $10 per month, including 2M characters of Simba 3.5 until 21 Nov 2026, 1.3M after1.3M characters of Simba 3.5, and Simba 3.5 on Starter is $15 per 1M characters, $10 until 21 Nov 2026. The older Simba 3.0 and 3.2 bill $10 per 1M characters on Starter, so the same balance buys 2M characters of them. Additional usage draws from your top-up balance. Free includes 500K characters each month and pauses when that balance runs out.
Compare all plansSpeechifyAI is Speechify's developer API, billed separately from the Speechify Reader app. The app that reads documents, articles, and books aloud is a separate product at speechify.com.
Visit the reading appGlobal Privacy Control honored. Your browser sent a Global Privacy Control signal, so we have opted you out of the sale and sharing of your personal information and switched off analytics and marketing cookies.