USD 
Swedish krona (SEK SEK)
Your cart is empty
You may check out all the available products and buy some in the shop
Return to shop-
Phonon-2 by Fermion Research: The Future of Efficient Speech Recognition
•
TL;DR Phonon-2 from Fermion Research is a 164 MB on-device automatic speech recognition model that hits 5.21% average word error rate across seven English benchmarks while running at 174x real time on an M5 MacBook Air. It turns high accuracy speech to text AI into a local capability on everyday…
-
Bespoke Nimble 9B: Local AI Classification and Decision Model
•
TL;DR Bespoke Nimble 9B is a 9 billion parameter typed decision model that replaces verbose text generation with structured probability outputs for enterprise classification, intent routing, content moderation, and fact verification. Built as a LoRA adapter on Qwen3.5 9B with an 8,192 token context and up to 255 choice options…
-
CLM v0.1 8B: The Faster Alternative to Generative AI for Agent Decisions
•
TL;DR CLM v0.1 8B replaces token generation with contrastive scoring for bounded decision tasks, delivering up to 9x lower latency than generative alternatives on tool routing, verification, and action selection workloads. Built on a frozen Qwen3 8B encoder with lightweight projection heads, it is Apache 2.0 open weights and optimized…
-
Confucius4-R2T2: The Voice AI Foundation Built for Live Captions and Agents
•
Confucius4-R2T2 is an open source streaming ASR model from NetEase Youdao that transcribes speech in real time with ultra low latency while never revising text once committed. Built on Qwen3-ASR with Longest Stable Prefix learning, it enables append-only output critical for live captioning, voice agents, simultaneous translation, and downstream NLP…
-
ZDTaichu5.0-9B: Spatial Reasoning AI That Runs at the Edge
•
TL;DR ZDTaichu5.0-9B is a 9 billion parameter multimodal vision language model engineered for spatial reasoning AI and embodied AI workloads that run on edge hardware rather than cloud clusters. Built on a Qwen3.5 language backbone with an NVIDIA C-RADIOv4 vision encoder, ZDTaichu5.0-9B excels at 3D scene understanding, multi view association,…
-
TeleOCR: One Model for Every Document, From Perfect PDFs to Crumpled Photos
•
TeleOCR is a lightweight, open source vision language model that parses both digital PDFs and camera captured document images inside a single framework. It delivers enterprise grade accuracy for text, tables, formulas, and complex layouts while remaining small enough for practical on premises deployment.
-
Gemini 3.8 Flash TTS: Features, Use Cases and Best Practices
•
TL;DR Gemini 3.8 Flash TTS is Google’s newest text to speech model, engineered for studio grade voice synthesis with expressive acting, custom voice creation, and multi speaker dialogue across more than 130 languages. Enterprises can use Gemini 3.8 Flash TTS to produce audiobooks, interactive voice agents, gaming dialogue, and branded…
-
Xiaomi MiMo V2.6 Flash RL: The Omnimodal AI Model Reshaping Enterprise Agents
•
MiMo V2.6 Flash RL combines a sparse Mixture of Experts architecture with large scale reinforcement learning to power self improving enterprise agents across text, image, video, and audio inputs.
-
Xing4.0-29B-A4B: China Telecom’s Agent Model That Runs on a Single GPU
•
Xing4.0-29B-A4B is a 29 billion parameter mixture of experts language model from China Telecom that activates only 4 billion parameters per token, supports up to 512K context length, and is optimized for autonomous agent workflows on single consumer GPUs.
-
Lyria 3.5: Google DeepMind AI Music Generation Guide
•
TL;DR Lyria 3.5 is Google DeepMind’s most advanced AI music generation model, capable of creating full-length songs with structured verses, choruses, and bridges from simple text prompts or images. It delivers high-fidelity 44.1 kHz stereo audio with improved vocals, lyrics, and musicality. Now integrated into the Gemini app and API,…
