Pins matching tag:"MIT License"

Microsoft introduces Phi-4, a 14B-parameter small language model trained mostly on synthetic data that scores 91.8 on November 2024 AMC 10/12 math tests, ahead of Gemini Pro 1.5 at 89.8, on Azure AI Foundry; open MIT-licensed weights follow on 8 January 2025.

Zhipu AI (Z.ai) releases GLM-4.5 (355B, 32B active) and GLM-4.5-Air (106B, 12B active), MIT-licensed hybrid-reasoning models for agents that rank third and sixth across 12 benchmarks, with API prices as low as $0.20 per million input tokens.

Z.ai releases GLM-4.5V, an open-weight 106B vision-language mixture-of-experts model with 12B active parameters, built on GLM-4.5-Air with a Thinking Mode switch and GUI-agent and grounding skills, and claims state-of-the-art results among similar-size open models on 42 public benchmarks.

Z.ai (Zhipu AI) releases GLM-4.6, its flagship coding model with a 200K-token context window, which wins 48.6% of head-to-head real-world coding tasks against Claude Sonnet 4 and uses over 30% fewer tokens than GLM-4.5.

Z.ai releases GLM-4.6V, a 106B vision-language mixture-of-experts model, and the 9B GLM-4.6V-Flash under the MIT license, the first GLM-V models with native function calling, trained for a 128K-token context and priced at $0.30 per million input tokens.

Z.ai (Zhipu AI) releases GLM-4.7, an MIT-licensed coding-focused flagship that scores 73.8 on SWE-bench Verified and 87.4 on τ²-Bench, ranking first among open models on Code Arena at the end of 2025.

Z.ai (Zhipu AI) releases GLM-5, a 744B-parameter (40B active) MIT-licensed model for long-horizon agentic engineering that scores 77.8 on SWE-bench Verified and became the first open-weight model to reach 50 on the Artificial Analysis index.

Z.ai (Zhipu AI) releases GLM-5.1 as MIT-licensed open weights, a 754B-parameter model built to work autonomously on one task for up to 8 hours, topping SWE-Bench Pro at 58.4.

Z.ai (Zhipu AI) releases GLM-5.2, an MIT-licensed 753B-parameter flagship built for long-horizon coding with a 1M-token context window, scoring 62.1 on SWE-bench Pro and 81.0 on Terminal Bench 2.1.