GPT 5 Upgrade to the Moveworks Reasoning Engine

Post updated with new information on Dec 12, 2025

The GPT-5 series of large language models stands out for delivering state-of-the-art reasoning. Better reasoning (i.e. thinking and decision making) represents a major leap for agentic tasks, offering improved instruction following, more effective plugin (tool) use. 

OpenAI released GPT-5.1 on November 13. We’re excited to share we’ve conducted a full set of evaluations to assess its performance relative to GPT-5 and GPT-4.1.The GPT-5 series of large language models stands out for delivering state-of-the-art reasoning. Better reasoning (i.e. thinking and decision making) represents a major leap for agentic tasks, offering improved instruction following, more effective plugin (tool) use.

Live today, in the Moveworks AI Assistant

Given the GPT 5.1's strong performance, we have upgraded to this new model in the Moveworks Reasoning Engine. No setup work or downtime is required to see this change, as this has already been made in your AI Assistant.
 

NOTE EXCEPTIONS: This upgrade applies to most Moveworks customers except GovCloud or Azure-only Orgs. Please contact your CSM for more information.

What are the benefits with GPT-5.1 in the Moveworks AI Assistant?

• As a reasoning model, GPT5.1 has a lot of the similar advantages of GPT5. 
  • Calls more relevant plugins (tools) in parallel to solve user requests.
  • More likely to fully explore solutions before escalating requests to other avenues for help.
  • Better summarization across more information. Responses are longer and cites more resources to help you find the right answer.
• We also saw that GPT5.1 relative to GPT5 had better latency. We can use more of the  reasoning capabilities without increasing wait time for users.

Beyond these above immediate benefits, GPT-5.1 lays the groundwork for these new capabilities:

• Personalized Responses. This tailors answers from the AI Assistant to each employee—improving relevance, clarity, and trust. By bringing key user attributes, including any custom attributes, into working memory, the AI Assistant can provide a more tailored response. Learn more here.
• There are upcoming improvements like richer reasoning steps to help you understand how the Assistant thinks, streaming of responses as they appear, ability to choose to use GPT-5.1’s high thinking levels. Stay tuned for these updates!

When will GPT 5.1 be available in Moveworks?

We have already started rolling out GPT5.1 to some customers. The GA dates are below:

🚧 Is there anything I need to do for this rollout?

This change has been made in your AI Assistant, no additional setup work or downtime work is required to see this upgrade.

NOTE: We do not offer official support for Optional Slots in Agent Studio. Customers have hacked different ways to make this work. For example, describing the slot like this: “The localized date on which this task is due, or null if the task has no due date.” 

GPT-5.1 may treat these types of directions differently. It may rather clarify with users to fill this slot; or decide to fill the slot value with a different value than before (empty string instead or “null”); or in some rare cases, it may not trigger a plugin. Please visit our Help Site for guidance for how to implement optional slots.
 

💡 Why does GPT-5.1 upgrade matter?

Model upgrades are a key part of how we keep the AI Assistant on the leading edge. Every upgrade is rigorously tested before rollout, and GPT‑5.1 has shown consistent improvements across core assistant tasks. 

You can expect upgrades like this a few times per year as part of our normal development cycle. These transitions happen automatically, and generally no action is needed on your end.

🛑 Are there any known issues?

• 12/10/2025 update: as we progressed with rollout, we identified an issue where the model’s internal reasoning or decision making progress is revealed in its response. This has been addressed as of Dec 10, 2025. This content is meant to stay behind the scenes because it is unnecessary or too technical. This reasoning is given as a prelude to its final response, so is not getting in the way of the end user receiving a helpful response from the Assistant. This was an intermittent issue on both Azure OpenAI and OpenAI. 
• With GPT5, the Assistant reasons through each user request more before response. This means responses might take a bit longer, but this also means users will notice an improvement in responses. We see the latency is more pronounced when using QuickGPT. 

GPT-5.2 is now available. When will Moveworks upgrade to the newest model? 

As an OpenAI partner, Moveworks received early access to GPT-5.2, and we’ve immediately put it to the test on our full agentic evaluation stack. When evaluating new flagship models at Moveworks, we measure not just how it scores on classic benchmarks but how well it actually completes real-world workflows—by testing its reasoning, tool use, policy adherence, and ability to orchestrate complex actions from start to finish. Read more about our early findings as well as expected release timelines in our latest blog post.