GPT-5.5 now available in Vincent

Today, alongside OpenAI’s launch of GPT-5.5, we are upgrading every Vincent and Clio Work customer to the new model for agentic work and document drafting. The upgrade sets a new bar for what Vincent can do on substantive legal tasks, and is the foundation for the next wave of autonomous legal capability we are building across our product line. 

Why we’re upgrading

We evaluated GPT-5.5 through Clio’s core evaluation set, designed by legal experts to reflect the work our customers do every day. This includes legal research, document analysis, drafting, discovery, and scenario-based advisory. These evaluations measure end-to-end system performance, including core agentic capabilities that support longer-horizon work, such as memory management and context retention across multi-turn sessions. They assess how effectively models orchestrated through Clio’s AI retrieve, integrate, and act on firm context, while leveraging our authoritative legal content library, one of the largest in the world.

We evaluated GPT-5.5 against every other frontier model available to us, including OpenAI’s prior-generation GPT-5.4 and models from other leading AI labs. This evaluation covered hundreds of scenarios and thousands of graded criteria, at multiple reasoning-effort levels for every model. When integrated into Vincent’s full system, including our agentic orchestration and legal data, GPT-5.5 delivered the highest overall score we recorded at 87.2%, outperforming every other frontier model we tested.

When powered by GPT-5.5, Vincent delivered the top overall benchmark score we recorded at 87.2%, higher than any other frontier model we tested.

Bigger gains on the hardest legal work

Two categories of work push frontier models hardest in our evaluation, and they are where GPT-5.5 moves furthest from the prior generation.

The first is legal research that requires citing the controlling authority, including the specific case, the exact statutory section, and the leading commentary, rather than merely describing the rule. On these tasks, GPT-5.5 delivers a roughly 20% relative improvement over the prior generation, closing gaps that earlier systems consistently left open.

The second is difficult, open-ended document work, which includes contract analysis, deal-point extraction, multi-document review, and discovery across large file sets. Earlier models would reliably surface the right answer but could miss the qualifying language, scope clauses, and secondary requirements that might alter its legal meaning. GPT-5.5 reads further into the document and captures key information: the survival periods, the fraud carve-outs, the jurisdictional conditions, the conditions of exercise. Across our document-analysis scenarios, this translates to a ~7% relative improvement over the prior generation, and the difference is even larger at higher reasoning effort. The result is an answer that is not merely directionally correct but more legally complete.

More efficient use of the context window

GPT-5.5 is markedly more efficient in how it uses tokens during reasoning. It spends fewer tokens deliberating internally for the same quality of answer than other frontier models we tested. In one comparison, it used ten times fewer reasoning tokens per tool call. In practice, this means two concrete things for our customers: faster responses, and more headroom in the context window for Vincent to retain context in long, multi-turn sessions and autonomous agent work.

What this means for our customers

Vincent’s agent mode is now powered by GPT-5.5, and customers will start to see the difference in their day-to-day work. This includes:

• Legal and matter context are seamlessly integrated. Vincent brings the relevant documents, notes, and matter history into its reasoning without user hand-holding.
• Drafted documents find more relevant precedent. Vincent’s search and retrieval work is more thorough, and that thoroughness shows up in the documents it produces.
• Routing is faster, deep analysis is better. Simple legal research questions are answered quickly; Vincent still triggers deep analysis when the task demands it and the quality of that deep analysis is meaningfully higher.
• Reasoning across provisions is stronger. On tasks that require connecting the dots between multiple contractual provisions or several authorities, Vincent delivers more complete and confident answers, with minimal boilerplate and unnecessary qualification.

Intelligence grounded in legal context

Vincent is built on top of Clio Library, a large, structured corpus of authoritative legal content spanning case law, statutes, and commentary across jurisdictions. This provides a high-quality retrieval layer that grounds every task in primary and secondary sources.

Outputs are generated against retrieved context, not prior probabilities. When Vincent cites an authority, it is drawn directly from retrieved sources used in its reasoning. When it analyzes a document, it operates on the full text. This shifts the system from plausibility to traceability.

As model capability increases, the value of that grounding compounds. GPT-5.5 handles larger context windows and longer reasoning chains more reliably, which means more of the retrieved signal is actually used. The result is tighter research, higher fidelity drafts, and less post-processing required to reach a usable output.

A foundation for the next generation

This intelligence upgrade will be foundational for the next generation of Clio’s agentic AI capabilities. Vincent will have the capability to autonomously locate and use matter context required for a task through integrated DMS tools without user guided manual selection. We are pushing the frontier of these agentic capabilities so Clio’s AI takes a more active role across legal work.

In Vincent, GPT-5.5’s extended autonomy and reasoning ability directly supports our roadmap of highly reliable autonomous legal agents that perform relevant legal work at scale. And, as we continue to deepen Vincent’s ability to leverage Clio Library and DocketAlarm data, the model’s stronger reasoning translates directly into more actionable guidance for legal professionals.

Looking forward

Consistency, reliability, and verifiability are the characteristics that unlock the full value of legal AI. That is the bar. Clio’s AI is built to meet it by integrating frontier models with our own systems, grounded in deep legal context. We partner closely with OpenAI and leading AI labs to push capabilities forward and bring the best of what’s possible to our platform, so those advances translate directly into the quality and scope of work our customers can achieve.