Z ai releases GLM-5.2 as top-ranked open weights model on Artificial Analysis index
Z ai has released GLM-5.2, which now holds the top position among open weights models on the Artificial Analysis Intelligence Index.
1. Performance and Benchmarking
Z ai’s GLM-5.2 has emerged as the leading open weights model on the Artificial Analysis Intelligence Index v4.1, achieving a score of 51. This performance surpasses other notable models such as MiniMax-M3 and DeepSeek V4 Pro, both of which scored 44. The model demonstrates significant improvements in scientific reasoning and technical evaluations, including gains in CritPt, HLE, and TerminalBench v2.1. Additionally, GLM-5.2 leads all open weights models on the GDPval-AA v2 benchmark with a score of 1524, placing it in a competitive position with proprietary models like GPT-5.5.
2. Technical Specifications and Availability
The model maintains the same architecture as its predecessor, GLM-5.1, featuring 744B total parameters and 40B active parameters. However, it offers an expanded context window of 1 million tokens, a substantial increase from the 200K limit of the previous version. Released under an MIT license, GLM-5.2 is accessible through Z ai’s first-party API and is supported by various third-party providers, including DeepInfra, Fireworks, and Baseten. While the model shows high intelligence, it also exhibits a higher output token usage per task compared to its peers, utilizing 43k tokens per task.
3. Cost and Market Positioning
On the Intelligence vs. Cost per Task Pareto frontier, GLM-5.2 is positioned as a highly competitive option, offering the lowest cost per task among models within its intelligence tier. The model is priced at $1.40 per 1M input tokens, $4.40 per 1M output tokens, and $0.26 per 1M cache hit tokens. Despite its increased output token consumption, the model's overall efficiency and performance metrics have established it as a significant development in the landscape of open weights artificial intelligence.