Top News

Z.ai’s GLM-5.2 open weights model with 753 billion parameters released on June 16, which outperforms OpenAI’s GPT-5.5 in several coding problems and is about one-sixth of the cost (in dollars) is a major achievement in open-source AI.

In VentureBeat, GLM-5.2 scored 62.1 on the SWE-bench Pro benchmark to do real-world software engineering tasks, significantly higher than GPT-5.5’s 58.6. FrontierSWE measures long-horizon task performance, GLM-5.2 scored 74.4 percent, far surpassing GPT-5.5 at 72.6 percent and Anthropic’s Claude Opus 4.8 at 75.1 percent.

The cost efficiency is also remarkable; the GLM-5.2 API is priced at $1.40 per million input tokens and $4.40 per million output tokens, or $5.80 per million. GPT-5.5 is priced at $5 and $30 per million input and output, so it is an additional $35 per million, so GLM-5.2 is about one-sixth cheaper than GPT-5.5, and offers comparable or better coding capabilities.

Architecture and Technical Innovation

With 753 billion parameters in the model, it has a huge optimization called IndexShare, which uses the same indexer in every four sparse attention layers. This reduces the number of compute operations per token by 2.9 at a maximum context length of 1 million tokens, as discussed in the VentureBeat article.

GLM-5.2 also features a more powerful Multi-Token Prediction layer to support speculative decoding and enables users to extend the length of the tokens during inference by up to 20 percent. The model also provides flexible “thinking modes” in which users can switch from Max effort for optimal logical problem solving to High effort for optimal performance and efficient token usage in terms of latency.

Open-Source Release and Market Impact

GLM-5.2 has been released under an unrestricted MIT open-source license by Z.ai, allowing organizations to freely download it from Hugging Face, customize it as needed, and run it locally or in virtual environments at very low cost of computing resources and electricity consumption, in contrast to many dual-use licenses that have strict governance policies.

This release comes at a time when American proprietary models are under the Trump Administration’s export control order banning foreign nationals from using Anthropic’s Claude Fable 5 model. For enterprise leaders, GLM-5.2 allows them to host cutting-edge AI solutions locally, less constrained by geographical proximity or commercial limitations.

The response from developers has been overwhelmingly positive from the moment of the launch day: Cline IDE reported that GLM-5.2 is “the first open weights model to score over 80% on Terminal-Bench,” scoring 81 on this benchmark and outperforming all other open models at the time. Kilo Code confirmed the integration with a lot of integration on launch with the 1 million token context window and Max effort mode in place in Kilo Code as well.

Z.ai also introduced the GLM Coding Plan for its use to implement the model at a low price of $12.60 per month for Lite, $50.40 per month for Pro, and $112 per month for Max (which is annually charged), with support for third-party coding tools such as Claude Code, OpenClaw, Cline Code, and Kilo Code.

In the performance measures of different benchmarks, GLM-5.2 scored 77 on MCP-Atlas (tool usage) over GPT-5.5’s score of 75.3; it also scored 54.7 on Humanity’s Last Exam with tool functionality enabled compared to GPT-5.5’s score of 52.2. In fact, it took first place in the crowdsourced Design Arena benchmark with an ELO rating of 1360, which was higher than Claude Fable 5.

The post Z.ai Iaunches GLM 5.2 appeared first on NewsX.