Urgent.News

What's breaking now, across thousands of outlets.

Sunday, 6 September 2026 Reading Room

AI

Microsoft เปิดตัว MAI-Transcribe-2 ถอดเสียง 60 ภาษา ภาษาไทยแม่นสุด ราคาถูกสุดในตลาด

Microsoft เปิดตัว MAI-Transcribe-2 ถอดเสียง 60 ภาษา ภาษาไทยแม่นสุด ราคาถูกสุดในตลาด Microsoft AI เปิดตัว MAI-Transcribe-2 โมเดลถอดเสียงพูดเป็นข้อความรุ่นล่าสุดเมื่อวันที่ 3 กันยายน 2026 โดยระบุว่าเป็นโมเดลถอดเสียงที่เร็วที่สุด แม่นยำที่สุด และถูกที่สุดในตลาดขณะนี้ รองรับ 60 ภาษา รวมถึงภาษาไทย และทำคะแนนอันดับหนึ่งบนมาตรฐาน FLEURS ด้วยอัตราความผิดพลาดของคำเฉลี่ย 5.2% ทั่วทั้ง 60 ภาษา [1]…

Original Thai Read in English

dev.to ·

Microsoft has unveiled MAI-Transcribe-2, a speech-to-text model capable of transcribing 60 languages, with Thai being the most accurate. Released on September 3, 2026, MAI-Transcribe-2 is the fastest, most accurate, and cheapest model in the market, supporting speaker diarization, word-level timestamps, keyword biasing, and code-switching.

Its performance outperforms competitors such as Google's Gemini 3.5 Transcribe, Gemini 3.1 Pro, OpenAI's GPT-transcribe, Scribe v2, and Whisper v3-large. MAI-Transcribe-2 is priced at $0.10 per hour, significantly cheaper than its predecessor at $0.36 per hour. The model aims to make speech-to-text services more affordable and efficient for businesses, particularly in call centers and telecommunications.

Brief written by urgent.news from Dev.to's own syndicated text. Machine-written — may contain errors; check the original before relying on it.

Read the original at dev.to →

More in AI

handoff: Give the Next AI Agent the Context It Actually Needs

AI-assisted engineering does not usually fail because an agent forgets everything. It fails because a new session receives the wrong kind of context.

  • Handoff skill creates compact document for AI agents to continue work
  • Includes live threads, references durable artifacts, saves to temporary directory
  • Redacts sensitive information while providing needed context

·dev.to

Agentic AI Development with Kiro: The Hidden DevSecOps Layer — Closing the Loop

Level 300 Some time ago we created a blog showing the capabilities of AI DLC and SDD to create quick and efficient prototypes as a MVP, the results were amazing, however, we omitted something…

  • Kiro bridges gap between prototypes and production-ready solutions in DevSecOps
  • Critical questions on security vulnerabilities, attack resistance, code quality remain unanswered
  • Development team fixed 344 issues using Kiro, showcasing AI-assisted DevSecOps potential

·dev.to

OmniVoice TTS model 600+ languages, voice cloning from 3-second clips, free open source.

OmniVoice TTS model 600+ languages, clone voices from 3-second clips, free open source. The k2-fsa team launched OmniVoice, a zero-shot text-to-speech model that supports more than 600 languages…

  • OmniVoice TTS model generates speech in over 600 languages
  • Developed by k2-fsa with diffusion language model architecture
  • Supports voice cloning and design from 3-second video clips

·dev.to

Hermes Command Cheat Sheet Update — Essential Commands to Know and Use

Hermes Command Cheat Sheet Update — Essential Commands for Practical Use by Nokka (Bird-Ga) | September 2026 This article was written by AI (DeepSeek V4 Pro) through Hermes Agent — reviewed and edited…

  • /steer (adjusting agent direction)
  • /bg (run tasks in the background)
  • /cron (repeat regardless of chat closure)

·dev.to

GPT-6 Astra: What’s Actually New in OpenAI’s New Frontier Model

OpenAI has released GPT-6 Astra, its newest frontier model, less than a week after Anthropic’s Claude Fable 5.1. OpenAI calls Astra the world’s most intelligent and aligned model yet.

  • GPT-6 Astra can independently complete tasks and generate documents without drafts.
  • Model excels at interacting with computers for form filling, CRM updates, and quality assurance.
  • Astra achieves Critical cybersecurity readiness, limiting exploit creation access to select users.

·dev.to

More from Sunday 6 September →