AI News
ElevenLabs Launches Scribe v3 and Scribe v3 Live
The new speech-to-text models bring claimed accuracy gains across 102 languages, a simple API upgrade and deployment choices that deserve a closer…
Jim Clyde Monge7 minZeniteq is now available on mobile for iPhone, iPad and Apple Watch.Download nowLearn more
All topics
Jim Clyde Monge8 min
Kim Monge9 min
Jim Clyde Monge8 min
Kim Monge9 min
Jim Clyde Monge8 min
James Rivera7 min
James Rivera8 min
Kim Monge7 min
James Rivera8 min
Jim Clyde Monge8 min
James Rivera7 min
Kim Monge8 min
Kim Monge8 min
James Rivera8 min
James Rivera9 min
Kim Monge8 min
Kim Monge10 min
James Rivera7 min
Kim Monge8 min
Kim Monge10 min
Jim Clyde Monge11 min
model release
Related reporting, guides, and analysis from Zeniteq.
AI News
Alibaba’s Qwen4-Max Preview Has a November 25 Deadline
Qwen4-Max-Preview brings a million-token context window, but developers need a Singapore-region Token Plan, separate credentials, and an exit plan before retirement.
Jim Clyde Monge8 minAI News
Gemini Robotics 2.5 Launches With a Partner-Only Rollout
Google DeepMind reports faster, more dexterous robot performance and a Mercedes-Benz factory pilot, but access remains limited to robotics partners.
Kim Monge9 minAI News
Meta Launches Muse Ember 1.0 for Low-Cost Agents
Meta’s new worker model combines multimodal input and a million-token context window, but its cost advantage depends on caching, output length and…
Jim Clyde Monge8 minAI News
Google Launches Gemini 4 Carbon With 1M-Token Decodes
Carbon’s November 10 broad launch brings paid API access, extended generation and temporary Argon-level pricing, with several migration details developers need to…
Kim Monge9 minAI News
NVIDIA Announces GR00T 2, Code Due in December
NVIDIA’s new world-action model predicts future video alongside robot actions, with reported gains in skill transfer and few-shot learning ahead of a…
Jim Clyde Monge8 minAI News
Seedream 5.5 Adds 100-Round Editing and Transparency
ByteDance’s new image model targets production rework with longer editing chains, native alpha-channel output and clearer text, while its quality claims remain…
James Rivera7 minAI News
Meta Launches Muse Ember for High-Volume Agent Work
Meta’s smallest Muse model targets repeated agent tasks, with low token prices, multimodal inputs and a long-context pricing threshold that changes the…
James Rivera8 minAI News
Nous Research Releases Arete-27B With Open Weights
The Qwen3.8 fine-tune gives local-model users downloadable weights and evaluation artifacts, but its political-behavior scores need a careful reading.
Kim Monge7 minAI News
Inkling 2 Small Launches on Tinker With Closed Weights
Thinking Machines’ 594-billion-parameter model is available for hosted use and fine-tuning, with Apache 2.0 weights planned after external safety testing.
James Rivera8 minAI News
OpenAI Launches GPT-6 Cyber With Critical Cyber Rating
OpenAI’s new security model promises stronger exploit development and tighter scope control, but access remains restricted to approved Daybreak Red customers.
Jim Clyde Monge8 minAI News
Inkling 2 Small Ships on Tinker With Weights Closed
Thinking Machines’ new model is available for hosted sampling and fine-tuning, while downloadable Apache 2.0 weights remain a conditional plan for next…
James Rivera7 minAI News
Google Gemini Omni 1.2 Flash Goes Generally Available
Google’s video model brings native 15-second clips, conversational editing and cheaper 720p generation to developers and businesses, while Gemini app access continues…
Kim Monge8 minAI News
Hydrogen-1.0 Brings Open Weights to Scientific AI
Deep Principle’s MIT-licensed research model pairs a million-token context window with hardware-specific builds, but its scientific performance still needs independent verification.
Kim Monge8 minAI News
42AI’s Behavioral World Model Targets Trader Decisions
42AI says brokerages already use its behavioral predictions, but its headline accuracy figure measures trading-flow ranges, leaving individual reliability and customer protections…
James Rivera8 minAI News
Baidu Open-Sources Unlimited OCR 2 for Long Documents
Baidu’s document-parsing model uses sparse visual references to keep each decoding step manageable as page counts grow.
James Rivera9 minAI News
Liquid AI Releases Open-Weight LFM3 Models
Six compact models share a common backbone, giving developers local options for text, vision, speech, multimodal assistants, and structured decisions.
Kim Monge8 minAI News
NVIDIA Nemotron 3.5 Transcribe: A Promptable Speech LLM
NVIDIA’s open-weight model handles transcription in 40 languages, with instructions controlling formatting, speaker labels, translation and other audio tasks.
Kim Monge10 minAI News
MiniCPM-OCR 1.0 and Jev Split Parsing From Review
OpenBMB and ModelBest’s two-model release separates document extraction from quality checking, giving builders a way to assess OCR results without replacing their…
James Rivera7 minAI News
ElevenLabs SFX v3 Improves Multi-Event Audio
ElevenLabs’ latest sound effects model promises better handling of multi-event prompts, stronger music and vocals, and cleaner audio across its creator tools,…
Kim Monge8 minAI News
Recursive Flint Offers Free Coding AI With 262K Context
Released on October 21, Flint gives coding-agent builders a free API endpoint, downloadable weights and tool support, with a few important limits…
Kim Monge10 minAI News
Baidu Releases Unlimited-OCR-2 for Long Documents
Baidu’s open document-parsing model uses sparse-reference attention to tackle long files, with local inference options and a new hundred-page benchmark.
Jim Clyde Monge11 min
