Paxa Labs (@PaxaLabs) header
Paxa Labs (@PaxaLabs) profile picture

Paxa Labs

@PaxaLabs

voice ai lab in bangkok. we make tts that's fast, cheap and actually sounds human. 50 voices in thai, english and mandarin 🇹🇭. contact: [email protected]

 Bangkok, Thailand.
paxalabs.com/
Joined October 2026
  • Tweets16
  • Following16
  • Followers18
11Media
Photo by Paxa Labs (@PaxaLabs)Photo by Paxa Labs (@PaxaLabs)Photo by Paxa Labs (@PaxaLabs)Photo by Paxa Labs (@PaxaLabs)Photo by Paxa Labs (@PaxaLabs)Photo by Paxa Labs (@PaxaLabs)Photo by Paxa Labs (@PaxaLabs)Photo by Paxa Labs (@PaxaLabs)Photo by Paxa Labs (@PaxaLabs)Photo by Paxa Labs (@PaxaLabs)
an AI voice that can break down mid-sentence. we're a 2-month-old voice lab and this is our flex 🔊
1
33
Paxa TTS Flash v2, voice: Cookie. try it free, no card → paxalabs.com
Speech and Language AI in English, Thai, and Mandarin | Paxa Labs

Speech and Language AI in English, Thai, and Mandarin | Paxa Labs

Paxa Labs is an AI research lab in Bangkok. Text to speech in English, Thai, and Mandarin, multilingual speech to text and OCR, and translation between Thai, English, Chinese, and other languages are...

paxalabs.com
10
She heard knocking at 3am. Headphones on 🎧
1
2
54
For everyone whose AI voiceover can't say Yaowarat 😤
1
1
1
3
138
Asked Foi Thong to whisper, get excited, laugh and sigh in one script. One take, no edits 👀.
40
We are announcing voice design for Paxa TTS Flash v2. Type "a woman in her fifties who reads ghost stories late at night" and hear her seconds later, with a name and a line in character. English, Thai, Mandarin. 5 credits per design. paxalabs.com/voice-design
1
31
We are announcing Paxa TTS Flash v2: Pixie. Our fastest, robust, expressive model yet. It whispers, laughs, cries, and changes emotion mid-sentence, on cue. 50 voices. English, Thai, Mandarin. First audio in as little as 90 ms. $10 per 1M characters.
1
2
2
54
more replies
The price stays at $10 per million characters, the rate of Paxa TTS Flash. To switch, set model to paxa-tts-flash-v2. The OpenAI-compatible /v1/audio/speech reads directions too. New accounts start with 100 free credits, about 10,000 characters.
Image from Paxa Labs (@PaxaLabs)
1
2
Hear every take and read the details in the launch post: paxalabs.com/blog/paxa-tts-f…
Paxa TTS Flash v2 (Pixie): AI Voices You Direct in Plain Words

Paxa TTS Flash v2 (Pixie): AI Voices You Direct in Plain Words

Paxa TTS Flash v2, nicknamed Pixie, performs the directions written into a line, such as a whisper or a laugh, in 50 voices that each speak English, Thai, and Mandarin, at the price of Paxa TTS Flash.

paxalabs.com
2
89ms to first audio. Fastest Thai voice in Asia, independently benchmarked by @speko_ai ⚡ We’re not done. V2 is coming: same speed, more natural, more accurate tones. Stay tuned 👀
Speko (YC S26) (@speko_ai) profile pictureSpeko (YC S26)
@speko_aiOct 13
The fastest Thai voice in Asia. Paxa Labs starts speaking in 89ms from Singapore. The next fastest takes 177ms. ElevenLabs takes 789ms. It's 2nd of 14 on getting the tones right, too. On Speko now. benchmarks.speko.ai
Image from Speko (YC S26) (@speko_ai)
1
2
45
Paxa Labs retweeted
Speko (YC S26) (@speko_ai) profile picture
Speko (YC S26)
@speko_aiOct 13
The fastest Thai voice in Asia. Paxa Labs starts speaking in 89ms from Singapore. The next fastest takes 177ms. ElevenLabs takes 789ms. It's 2nd of 14 on getting the tones right, too. On Speko now. benchmarks.speko.ai
Image from Speko (YC S26) (@speko_ai)
1
2
2
9
416
Our next generation model, Paxa TTS Flash v2: "Pixie", is coming soon. Truly SOTA TTS with top of the line naturalness and blazingly fast TTFT/latency for only $10/1M characters.
1
4
4
6
681
Load more