Prompt and Model

ModelsFrontier models and releases Open weight models Mistral Aleph Alpha Silo national initiatives text image video audio multimodal Small and on-device models Reasoning models Benchmarks and what they actually measure Model cards and documentation Release changelog Context windows, rate limits and pricing per token Apache-2.0 MIT Llama Community Gemma terms Quantisation and running models locally Performance in EU languages beyond English Deprecations and model retirement datesRegulationEU AI Act: risk tiers and timeline General purpose AI obligations High-risk system requirements Prohibited practices Transparency and labelling duties lawful basis training data DPIAs Copyright and text and data mining exceptions AI Act enforcement and national authorities Product liability and the AI Liability Directive finance health employment GPAI Code of Practice The EU AI Office and what it does Standardisation: CEN-CENELEC JTC 21 Member state implementation and national authorities Overlap with the Digital Services Act Data Act and Data Governance Act NIS2 security duties EDPB opinions on AI and personal data Enforcement actions and fines so farBuilding with AIPrompting and context design Retrieval augmented generation Fine tuning and adaptation Agents and tool use Evaluation and testing Guardrails and safety layers Cost and latency Self hosting vs API Vector databases and infrastructureIn practiceAI at work and workforce effects AI in the public sector AI in healthcare under EU rules AI in education Creative industries and consent Procurement of AI systems Deployment failures and what went wrong AI in banking and insurance under EU rules AI in recruitment and the high-risk classification AI in local government Works councils and employee consultation Insurance and indemnity for AI outputs Vendor lock-in and exit planningSafety & societyBias and discrimination Deepfakes and synthetic media Misinformation Privacy and surveillance Energy and water use of AI Labour behind the models Concentration and competition Research on capabilities and limits Model evaluations and red teaming Watermarking and C2PA content provenance Elections and political content rules Child safety obligations Copyright litigation and its outcomes Accessibility gains from AINewsCompanies and labsOpenAI Anthropic Google DeepMind Meta AI Mistral AI Aleph Alpha Stability AI Hugging Face Cohere DeepSeek Alibaba Qwen Black Forest Labs Silo AI NvidiaCompute and infrastructureGPUs and accelerators EU datacentre buildout LUMI Leonardo JUPITER OVHcloud Scaleway IONOS Grid capacity and energy contracts Chip export controls Inference cost and where it is falling Cooling, water use and siting

Quantisation and running models locally

Quantisation and running models locally - reporting from Prompt and Model.

No articles filed here yet. New reporting appears as it is published.

Prompt and Model covers Artificial Intelligence.

About · Editorial policy · Privacy · Contact · RSS

© 2026 promptandmodel.com. Independent publication. All articles are written from reported sources, which are linked on every page.