MiniMax
Speech 2.5 Turbo Preview Voice Clone
The text speech 2.5 model by MiniMax provides industry-leading zero-shot voice cloning. With sub-300ms latency and high-fidelity 48kHz output, it transforms text into natural speech with emotional cues like breaths and laughter instantly.

Modalities

Input: TextInput: Audio
Output: Audio

Input

Output

API Usage Examples
curl --request POST "https://gptproto.com/api/v3/minimax/speech-2.5-turbo-preview-voice-clone/text-to-audio" \
  --header "Authorization: Bearer $GPTPROTO_API_KEY" \
  --header "Content-Type: application/json" \
  --data '{
    "audio": "",
    "custom_voice_id": "",
    "text": "Hello! Welcome to GPTProto! This is a preview of your cloned voice. I hope you enjoy it!",
    "need_noise_reduction": false,
    "need_volume_normalization": false,
    "accuracy": 0.7
  }'
Speech 2.5 Turbo Preview Voice Clone pricing

Start from the cost of a single sample and pick a testing budget. GPTProto rates are 40% below list price.

GPTProto · Price est.

Estimated from the rate card. Final charges may vary.
40% off

No extra settings

Cost per request$0.5003$0.8338

Top up

GPTProto vs official pricing.
Requests
You pay
40% off
$100
You receive$100.00

Save$65.93 (40%)vs MiniMax official

Related Models
All Models
ModelResolutionInput → Output
Speech 2.5 Turbo Preview Voice CloneCurrent
$0.50 per time—
Input: TextInput: Audio
Output: Audio
Gemini 2.5 Flash Preview TTS
$0.30 / $6.00 per 1M—
Input: Text
Output: Audio
Speech 2.6 HD
— / $60.00 per 1M—
Input: Text
Output: Audio
GPT 4o Mini TTS
$0.42 / $8.40 per 1M—
Input: Text
Output: Audio
Gemini 2.5 Pro Preview TTS
$0.60 / $12.00 per 1M—
Input: Text
Output: Audio
Speech 2.5 Turbo Preview
— / $36.00 per 1M—
Input: Text
Output: Audio
Speech 02 Turbo
$0.002 per time—
Input: Text
Output: Audio

text speech 2.5 Key Features

Advanced features of the speech-2.5-turbo-preview-voice-clone model, optimized for high-fidelity text to audio tasks.

Zero-Shot Voice Cloning

Create a clone in seconds using just a 3-6s audio sample. No fine-tuning required for high-fidelity results.

Sub-300ms Low Latency

Optimized for real-time chat with a TTFA of ~280ms, making it faster than most industry competitors for live use.

Emotional Prosody & Tags

Native multimodal architecture generates non-verbal cues like laughter and breaths for truly human-like output.

Cross-Lingual Capabilities

Clone a voice in one language and have it speak another of the 25+ supported languages while keeping its accent.

text speech 2.5: Frequently Asked Questions

Get answers about text and speech capabilities of the 2.5 model. Learn about cloning, latency, and how to integrate this text-based AI into your application via the GPTProto.com platform.

How much text is needed for speech synthesis?

The speech-2.5-turbo-preview-voice-clone model can handle up to 10,000 characters of text per request. This allows for long-form content generation, though we recommend breaking extremely long text into chunks for optimal streaming performance. This ensures the text to audio conversion remains stable and responsive for the end user.

Does text speech 2.5 support zero-shot cloning?

Yes. You only need a 3-6 second audio reference. The model uses this sample to clone the timbre and prosody of the target voice immediately without any additional training or fine-tuning, making it ideal for rapid deployment. The resulting speech maintains the accent and unique characteristics of the original speaker with high accuracy.

What is the latency for text to speech tasks?

This 2.5 turbo model is engineered for real-time interaction. It achieves a Time-to-First-Audio (TTFA) of approximately 280ms. This sub-second latency ensures that your voice assistants respond to text inputs almost instantly, providing a fluid conversation experience that feels human and natural.

Can the model handle bilingual text inputs?

Absolutely. It excels at code-switching, meaning it can transition between two languages, such as English and Chinese, within a single text sentence without losing voice quality or creating audio glitches during the speech output. This makes it perfect for international applications and global users.

Is 48kHz high-fidelity audio supported?

Yes, the 2.5 model supports 48kHz sampling rates. This high-definition output is suitable for professional media production, gaming, and any application where the audio quality of the speech must meet broadcast standards. You can specify the response format as flac or pcm to preserve this high-fidelity quality.

How do I access text speech 2.5 on GPTProto?

You can access the model via our OpenAI-compatible API. This allows for unified billing and no minimum monthly spend. Simply update your base URL and use your GPTProto key to start generating high-quality speech from your text. Our platform provides the lowest latency routing to ensure your application performs at its best.

Related Articles

Guides, comparisons, and updates related to this model.

All Articles