#GLM 5.2
10 articles with this tag
Most read in #GLM 5.2, last 30 days

Crusoe splits RL so a 9B beats a 1T
Crusoe says Managed RL lifted a 9B Qwen from 53.1% to 73.9% on held-out FinQA, beating a 1T model.

Databricks pushes open weights as frontier parity
Databricks field engineers cite seven pressures, including blocked malware analysis and retention shifts, to argue enterprises should run open-weight models via Unity Gateway.

Fireworks Pitches One-Click Loop, Proof Is Thin
Fireworks AI pitches a train-to-deploy loop to replace closed models, but key speed and quality wins lack baselines and pricing.

Z.ai pushes GLM-5.2 open weights near frontier
Zhipu pitches GLM-5.2 as open weights near Opus 4.7 to 4.8 on agentic benchmarks, pushing on-premise and fine-tune adoption.

Hugging Face Engineer Automates Job with AI Agents
Niels Rogge from Hugging Face shares how he uses AI agents to automate his job, from outreach to researchers to improving model discoverability on the Hugging Face Hub.

AI Inference: 10x Faster Models & Self-Optimization
Philip Kiely and Ali Taha of Baseten discuss AI inference, LLM optimization, speculative decoding, and the engineering behind cutting-edge AI models.

AI Agent Solves Security CTF Cheaply
Crusoe Cloud's AI agent solved a security CTF in under 2 hours for ~$13, proving effective tooling matters more than just a frontier model for cost-effective AI agent tasks.

Local AI Models: The Desktop Frontier is Here
Ahmad Osman of Osmantic explores the rapid advancement of local AI, predicting powerful models on consumer hardware and advocating for sovereign AI.

Databricks Benchmarks AI Coding Tools
Databricks benchmarks AI coding agents on its multi-million line codebase, finding open-source models competitive and token price an unreliable cost indicator.

Z.AI's GLM 5.2 Sparks Chinese Open Source AI Moment
Z.AI's GLM 5.2, a powerful open-source AI model with a 1M token context, is making waves, challenging Western AI dominance and highlighting the growing Chinese AI sector.

