AI is no longer optional in commercial real estate, it's the new operational baseline

At a growing number of commercial real estate firms, the question is no longer whether to use AI, it's whether your team is using it well enough. Adventures in CRE, which maintains one of the industry's most-referenced AI tool directories, stated plainly in its Summer 2026 edition that regular AI use has become the expected baseline across acquisitions, development, management, investor relations, and brokerage. Professionals who treat it as optional are already operating at a disadvantage relative to peers who have embedded these tools into daily work.

That shift matters operationally because it changes how CRE teams should be evaluating vendors, allocating training budgets, and structuring workflows. AI is not a parallel track anymore, it is the track.

The model stack CRE professionals are actually using

Adventures in CRE now maintains a live AI model leaderboard calibrated specifically for CRE professionals, pulling performance data from Artificial Analysis and updated daily. The leaderboard uses the Artificial Analysis Intelligence Index, a composite quality score averaged across standardized benchmarks including MMLU-Pro, GPQA Diamond, MATH-500, and long-context reasoning tests. As of July 2026, Anthropic's Claude Opus 5 holds the top spot with an Intelligence Index score of 60.7, priced at $5 per million input tokens and $25 per million output tokens.

OpenAI's GPT-5.6 Sol ranks fourth on the leaderboard with an Intelligence Index of 58.9, priced at $5 per million input tokens and $30 per million output tokens, according to Adventures in CRE. For teams optimizing on raw speed rather than analytical depth, Google's Gemini 3.5 Flash leads throughput at 288 tokens per second. The spread between these models in cost and capability is meaningful for firms running high-volume underwriting or market analysis pipelines, where token costs compound quickly at scale.

Adventures in CRE also flags a value composite for its audience, noting that for CRE work, accuracy matters far more than speed, a meaningful framing for procurement teams evaluating API contracts or enterprise AI licenses. The site's methodology pulls release dates and model identifiers from OpenRouter and updates pricing from the Artificial Analysis API, giving operators a near-real-time view of the competitive model landscape without having to monitor multiple vendor dashboards themselves.

From workflow tool to market catalyst

Jack Mullen’s June 2026 Forbes Finance Council opinion addresses a different question: how AI companies affect demand for offices and data centers.

Mullen expects an uneven property recovery, rather than an industry-wide rebound driven by faster underwriting.

Bespoke tools built by non-technical teams

One of the more operationally significant observations from Adventures in CRE is that generative AI now allows non-technical CRE professionals to build custom solutions for specific problems quickly and at low cost. That capability closes a gap that previously required dedicated engineering resources or expensive third-party software development. A VP of acquisitions can now prototype a deal screening tool or a lease abstraction workflow without writing a line of code, then iterate on it based on actual deal experience.

Spencer Burton, a co-founder at CRE Agents, a firm building AI agents purpose-built for commercial real estate, and a co-founder of Adventures in CRE, has noted that an AI-first mindset frees teams from repetitive tasks and redirects attention toward strategic thinking, better decision-making, and relationship management. Those are precisely the activities that differentiate firms in a transaction-driven business. The emergence of specialized CRE agent platforms, alongside general-purpose models, means operators now face a real vendor evaluation decision: a horizontal model accessed via API, or a vertical application built for CRE workflows specifically.

What CRE operators should evaluate now

The practical implication for enterprise CRE operators is a vendor and training evaluation window that is closing fast. Adventures in CRE updates its tool directory at least quarterly and notes that the space is moving fast enough that tools are being added and removed based on adoption, continued development, and emergence of better alternatives. Waiting for the market to stabilize before committing to a stack is itself a strategic choice, and based on the trajectory described in both sources, it is likely a costly one.

For teams mapping their AI strategy, the Summer 2026 leaderboard data from Adventures in CRE gives a defensible starting point for model selection: Claude Opus 5 for maximum analytical depth, GPT-5.6 Sol as a competitive alternative, and Gemini 3.5 Flash where throughput is the priority. The next decision is whether to run those models directly or through a CRE-specific agent layer. That choice will shape what the team can actually build on top of them.

Adventures in CRE plans to continue updating its directory and leaderboard on at least a quarterly cadence, with major model releases triggering off-cycle updates, meaning the July 2026 rankings will likely shift again before year's end.