Latest AI and tech news
When new evidence kills one conclusion, do the decisions built on it change too? I tested six update strategies, then a local model....
TL;DR: Solace Agent Mesh decision nodes work with any model that speaks TypeSafe’s decision API, so they can run on hardware you control. We pointed the same workflow at two open-weight decision models on one Mac and measured them against Jev on 40 s...
Kristijonas Adomaitis has published Kadmos-1 on GitHub, a decoder-only transformer language model with 45.8 million parameters that holds a short conversation and writes small Python functions. The architecture, tokenizer and weights were built in th...
Epoch AI researchers say the widely available AI model GLM 5.3 turned public security fixes into working exploits before updates reached Firefox users. It autonomously compromised six Firefox releases for about $1,000 in API fees each without trigger...
Revolut’s David Abitbol says GLM-5.3 fits on 4 GPUs but runs short of memory. The reason is in vLLM’s source: every GPU keeps a full copy of each chat’s memory, the KV cache....
NVIDIA Nemotron 3.5 Transcribe 30B A4B is an open-weight, offline speech-and-audio model that turns recordings and text instructions into text. It supports transcription, translation, speaker-attributed and contextual transcription, timestamps, langu...
Edge0 has released Audio8-TTS-Realistic, an open-weight text-to-speech model for zero-shot voice cloning. Audio8-TTS-Realistic is a 0.6B-parameter model that clones a voice from 1 reference clip and transcript. It supports 23 languages, 12 more than ...
You may not need an expensive AI rig to run a serious local model....
A case for a Raspberry Pi 5 with the official AI HAT installed, plus a camera mount on top — an ‘all-in-one setup to tinker with the AI Module.’ Great fit for Adafruit’s Pi + machine-learning audience exploring vision pipelines or on-device AI. Compa...
Today we are taking a look at ASRock Industrial's highly portable Ryzen AI Max+ 395 system, the AI Box-A395. The barebones tower system packs a punch, as well as a handle to quickly pack up the PC...
On Thursday, Mistral released the weights for its frontier open model, Mistral Large 4 (ML4), on Hugging Face....
Nemotron 3.5 Transcribe is NVIDIA's open-weight speech-to-text model for recordings. Specs, accuracy, 40 languages, license, and how to run it.
NVIDIA has released Nemotron 3.5 Transcribe, an open-weights speech LLM built for transcription. One model handles 40 languages and recordings up to 1 hour. A text instruction picks the output style for each request. Options include verbatim, written...
Cheap models give you bargaining power. A working output gives you leverage....
Saulo Verissimo has published whistle-esp32 on GitHub, a speech-to-text demo that runs entirely on an ESP32-S3. Hold a button, say something like “make it red” or “flash the light blue”, and the board’s onboard WS2812 LED does it. No network, no clou...
Nvidia on Wednesday launched Nemotron 3.5 Transcribe, its latest multilingual transcription model. The new open-weight model, the company says, ranks first on Hugging Face’s Open ASR open model leaderboard for English, with an average word error rate...
Survival LLM is a free, open source AI assistant and reference library that runs entirely on a local machine. Ask it how to keep food safe during a blackout, treat a burn, or purify water, and it answers from stored references and cites the articles ...
Ananthu Ashok, writing for Open Source For You, covers esp32-pie-asr, an open source project that runs live English speech recognition on an ESP32-P4 with no cloud service and no separate AI accelerator. Developer Boumedine Billal built it to display...
How I got OpenAI’s Whisper to transcribe on the user’s own device with onnxruntime-web, int8 weights and one CPU thread, and cut Urdu word…...
A 96 GB graphics card costs $16,999 used. A 96 GB Mac gives your AI only 72 GB. So what is the real price of 96 GB, and what does it let…...
How can a model get better and worse at the same time? The answer says a lot about where local AI is heading....
Why the cheapest path is a used server, what each hardware tier really costs, and which open models are worth loading first....
From 8GB MacBooks to 512GB Mac Studios, discover which local AI models actually fit, how fast they run, and which Mac is worth buying for…...