Enterprises are moving quickly from AI experimentation to production, and one thing is becoming clear: no single model is right for every workload. Some use cases need long-context reasoning. Others need cost-efficient summarization, coding support, domain-specific generation, guardrails, speech transcription, or technical language understanding. The ability to choose the right model for the right job is becoming central to both AI performance and total cost of ownership.
Today, we’re excited to announce that GLM 5.2 is now available on OCI Enterprise AI through Model Import.
Model Import in OCI Enterprise AI allows customers to bring supported open-source and third-party models into OCI, deploy them on dedicated AI clusters, create endpoints, and use them through OCI’s enterprise-ready AI platform.
Why GLM 5.2 matters
GLM 5.2 is Z.ai’s latest flagship model for long-horizon tasks, with a strong focus on extended-context reasoning, advanced coding, and agentic workflows. Z.ai describes GLM 5.2 as delivering long-horizon capabilities with a 1M-token context window, making it well suited for workloads that need to reason across large amounts of information.
That combination is especially relevant for enterprise use cases such as:
- Analyzing large document collections, contracts, logs, or technical manuals
- Assisting with software engineering, code review, debugging, and modernization
- Supporting agentic workflows that require planning across many steps
- Summarizing and reasoning over long-running conversations or operational histories
- Building domain-specific assistants that need deeper context and stronger task persistence
For customers building on OCI Enterprise AI, GLM 5.2 adds another powerful option for teams that need advanced reasoning and long-context capabilities while keeping deployment, governance, and infrastructure decisions within OCI.
Expanding model choice while improving TCO
The addition of GLM 5.2 is part of a broader effort to give customers more flexibility in how they build and scale AI applications on OCI.
Model choice is also a cost optimization strategy. Not every workload needs the largest or most expensive model. By giving teams access to a broader model portfolio with no application code changes needed, OCI helps customers right-size model selection based on latency, quality, throughput, domain fit, and cost requirements.
In June, we also added support for several additional models through Model Import, including:
- DeepSeek V4 Flash and DeepSeek V4 Pro
- Google MedGemma 27B Text IT
- MiniMax M3
- OpenAI Whisper Large V3 Turbo
- Additional Qwen models
Together, these models broaden the range of workloads customers can support on OCI Enterprise AI. Smaller and specialized models can help reduce cost for high-volume tasks. Larger reasoning and coding models can be reserved for more complex workflows. Speech, safety, medical, and domain-specific models can be used where fit-for-purpose performance matters most.
This model mix helps customers optimize for business outcomes rather than simply defaulting to a single model for every use case.
Getting started
To get started with GLM 5.2 through Model Import on OCI Enterprise AI:
- Open the OCI Console
- Import the model using one of these options:
- Create a hosting dedicated AI cluster for the imported model with a recommended unit shape.
- Create an endpoint.
- Call the model through OCI Generative AI API, SDK, or use the model in the playground.
GLM 5.2 gives customers another strong option for long-context reasoning, coding, and agentic AI workloads on OCI. Combined with the broader set of models added in June, OCI Enterprise AI continues to expand model choice while helping customers optimize performance and TCO for production AI.