Skip to content
OOpenSpeech

OPEN SPEECH. CLEAR PRICES.

Pay for the speech you create.

One credit equals $0.01. Buy credits up front and use them in the studio or API. No subscription, no automatic top-up, and no free generation allowance.

500 credits

$5.00

Buy credits →

2,000 credits

$20.00

Buy credits →

5,000 credits

$50.00

Buy credits →

19 models, priced before you press Generate

Each recording costs the greater of its minimum or its character-based price, rounded up to a whole credit. Example: 1,000 Kokoro characters cost 15 credits ($0.15). Shorter recordings have a 10-credit minimum.

ModelMinimumPer 1,000 charactersCharacter limit
Kokoro-82MOpen models10 credits15 credits2,000
Chatterbox TurboOpen models10 credits30 credits1,000
Orpheus TTSOpen models30 credits40 credits300
Qwen3-TTSOpen models10 credits30 credits1,000
ChatterboxOpen models10 credits30 credits1,000
Chatterbox MultilingualOpen models10 credits40 credits300
Sesame CSM 1BOpen models20 credits60 credits200
Parler TTSOpen models20 credits50 credits300
PhemeOpen models30 credits50 credits300
Inworld Realtime TTS 2Premium models5 credits20 credits2,000
Inworld TTS 1.5 MiniPremium models5 credits20 credits2,000
Inworld TTS 1.5 MaxPremium models5 credits20 credits2,000
MiniMax Speech 2.8 HDPremium models5 credits30 credits2,000
MiniMax Speech 2.8 TurboPremium models5 credits20 credits2,000
ElevenLabs v3Premium models5 credits50 credits2,000
ElevenLabs Flash v2.5Premium models5 credits25 credits2,000
ElevenLabs Turbo v2.5Premium models5 credits25 credits2,000
ElevenLabs Multilingual v2Premium models5 credits50 credits2,000
Gemini 3.1 Flash TTSPremium models15 credits30 credits1,000

Credits are reserved before the provider starts. Successful recordings spend that reservation. Failed or canceled provider jobs return credits automatically; a network timeout may take a few minutes to settle. Closing the browser does not cancel a submitted job.

Credit packs do not expire. Save your private workspace recovery key before purchasing: it restores access after browser data is cleared. Keep downloaded audio locally because provider hosting is temporary.

OpenSpeech prices include provider costs and service margin. They are not Replicate’s direct prices. Rates shown here apply to future generations, and the studio shows the exact credit charge for your script.

All models run through Replicate. Open and proprietary models are labeled separately. Preset English voices are available; the studio does not currently expose every upstream cloning or language feature.

Payments are processed by SuperTools through Stripe. Requests are processed one at a time per workspace, with model deadlines and daily service capacity limits. For unused-credit refund requests, use the merchant contact details on your payment receipt.