TokenPost topic
Local AI
15 TokenPost stories are tagged Local AI. The newest appear first.
Latest Local AI stories
- Tipster Says Apple Is Developing Harbor for Local AI on Macs
The claim describes open-weight models running on Apple silicon and clusters of two or more Macs linked by Thunderbolt 5.
· Technology
- Musk Pushes Back on Altman’s 5% DRAM-Use Estimate in Local AI Debate
OpenAI CEO Sam Altman argued cloud systems could use scarce memory more intensively, but did not define which computers his 5% estimate covered.
· Technology
- Vitalik Buterin Says Local AI Could Reduce Reliance on Crypto Web Apps
He said websites can identify visitors through device details and behavior, and described using an AI-generated command-line tool to change an IPFS hash tied to ENS.
· Technology
- NVIDIA Says 64GB DGX Spark Is Available Through Partners
The developer system starts at $4,999 and is offered by six manufacturers.
· Technology
- Liquid AI Debuts Six Open-Weight Models for Local AI Workloads
The family covers text, vision-language, audio and decision tasks, alongside an omni model that handles text, images, video and audio.
· Technology
- Block Unveils Dessa Speech-to-Speech Model for Local MacBook Use
The model was trained from scratch on human conversation audio and can run faster than real time on a MacBook Pro M4 Max or a higher-spec device.
· Technology
- NVIDIA Expands DGX Spark Clusters to Eight Systems With One Terabyte of Unified Memory
Eight systems can run Z.ai's 753-billion-parameter GLM 5.3 model, and benchmark results vary by model and network layout.
· Technology
- NVIDIA Adds Flint Model and Management Tools to Local AI Lineup
An October DGX Spark update adds a Wi-Fi Assistant and device troubleshooting guidance. The 64GB configuration starts at $4,999 and is scheduled to roll out Oct. 23.
· Technology
- Nvidia Will Offer 64GB DGX Spark Systems for $4,999 Starting Oct. 23
Two linked systems can share 128GB of memory and support AI models with up to 200 billion parameters.
· Technology
- PrismML Packs Its 9B Bonsai 2 Model Into 2GB for Local AI
The 2B, 4B and 9B releases use main weights limited to −1, 0 and +1, with file sizes from 0.43GB to 2GB.
· Technology
- Jailbroken PS5 Runs Qwen3.8-27B Locally at About 20 Tokens per Second
Both 27B benchmarks use roughly 2-bit quantization. Prompt processing is slower, and long inputs can delay a first response by several minutes.
· Technology
- AMD Launches Ryzen AI Halo Developer Platform With 192GB Memory
The compact system uses a Ryzen AI Max+ PRO 495 processor and can allocate up to 160GB of shared memory to its integrated GPU.
· Technology
- Microsoft Brings Local AI Inference to GitHub Copilot This Month
Developers will be able to switch automatically or manually between cloud-based and device-based models in GitHub Copilot.
· Technology
- Microsoft Sets Windows, Surface Event on Local AI and RTX Spark
The Oct. 7 event will cover the latest Windows and Surface developments as Microsoft works with Nvidia on local-AI hardware.
· Technology
- Ghost Sells Core Personal AI Computer for $3,499
The company says Core runs models and stored context locally and connects with users’ apps, files and devices.
· Technology