Track 3 · Economic World Model Challenge · opens December 2026

A city plan
you can run.

Agentic City is an agent-native planning studio for Hong Kong's Northern Metropolis. A plan is built, priced, checked and submitted here, not drawn on a slide.

planning area
24.3 km²
planning zones
289
existing buildings
20,924
fixed budget
HK$ 10 bn
target year
2035

Live illustration drawn from the published plan. The traffic is not simulation output.

01 / THE IDEA

The idea in four lines.

  1. 01

    The city is a state machine, not a picture.

    Zoning, permitted uses, buildable land, demolition and pond-filling rules, the road graph, capacity and prices are machine-readable. Every operation is checked, priced, versioned and undoable, and the 2D plan and 3D city are two views of one plan.

  2. 02

    The agent is a colleague, not a chat box.

    You describe an intent; the assistant turns it into a concrete proposal and its cost, then stops. Nothing is built until you approve it. Spending is the boundary between conversation and consequence.

  3. 03

    The plan is meant to be run.

    A submission is machine-readable, and the committee loads it into an Economic World Model where households, firms, investors and service organisations live inside it.

  4. 04

    Two kinds of agent.

    The planning agent helps you operate the city. The city agent lives in it. The first makes the tool usable; the second is what makes the city agentic.

02 / THE FILM

The idea, in 94 seconds.

A short film cut from the project's own recordings: why a plan should be a machine, why the agent stops before it spends, and what it means to run a city plan.

1:34 · narrated in English or Chinese, captions in the picture · subtitle files: EN .srt · 中文 .srt

03 / THE CONTRACT

Nothing is built
until you approve.

The agent proposes, the human approves, the core executes. The whole product rests on that contract, and it is why a language model can be given write access to a city at all. Below is the demo's fire-services scenario, replayed as it was recorded. The click is yours.

Agentic City · Planning studio Saved revision 0
AVAILABLE PLANNING BUDGETHK$ 10,000,000,000.00
2D planning mapLand, buildings and transport connections
The planning district, as shown on the studio's 2D map Zone Z6-150 with a proposed fire station outlined Zone Z6-150 with fire station 1 built The district with the new fire stations labelled A proposed 146 metre access road, dashed The 146 metre access road, built
Fire services · Step 1 of 4 · Build three stations
Planning AgentCITY DATA · RULES · TOOLS

A replay of the demo's guided fire-services scenario: the instructions, prices and travel times are the demo's own. To drive the real studio: open the live demo ↗

04 / THE EVIDENCE

What the demonstration establishes.

Fire services

4.92→3.44min

mean response −30%

Names three eligible sites, approves construction, later proposes one ~146 m access road.

Residential

46,576residents

+660 resident capacity

Picks a site, demolishes a building, fills a pond, builds six blocks.

School

450places

9.33 min mean journey

Sites a campus on buildable land, assigns three residential origins.

None of these numbers is typed in: the capacity and travel figures are computed on the actual V6 network. In the public demo they are replayed from a pre-validated script; see Scope.

05 / THE LOOP

Plan, simulate,
observe, iterate.

The loop is the point. Assessment is not a verdict at the end; it is the signal that drives the next revision.

  1. Read the briefboundary · land use · rules · budget
  2. Plan with your agentbuildings · roads · policies
  3. Self-assesscapacity · access · services · costiterate
  4. Submitspatial_plan.zip + policies.yaml
  5. Committee checksformat · consistency · rules · budget
  6. EWM scenario simulationhouseholds · firms · investors · services
  7. Agent satisfaction + expert reviewnovelty · soundness · feasibility · social value

06 / THE STUDIO

Twenty-eight chapters of the studio at work.

Two films, cut into chapters. Pick the question you care about. The clips stream from the repository's attachments on GitHub.

1-01 · 0:14 · 英文旁白

What Agentic City is

    07 / UNDER THE HOOD

    Four layers on one shared city record.

    The property that matters is that nothing is stored twice: the 2D plan, the 3D city, the assessment panel and the agent conversation all read and write the same plan state.

    1. InterfaceOne workspace, six views: the planning studio with its 2D map and 3D city, the evaluation panel, the policy workspace, the competition guide, the video guides, and model settings. 2D and 3D are two views of one revision.
    2. AgentTurns intent into a concrete proposal, and then stops: the participant sees what will be built and what it costs, and only after approval does the city change.
    3. Planning coreThe operations a planner can actually perform. Each one is validated against the city model and priced, and the price comes out of a single budget shared by construction, demolition, land preparation, road building and policy commitments. State changes are written as revisions.
    4. City modelThe compiled 2035 baseline for a 24.3 km² slice of the Northern Metropolis: 20,924 existing buildings, planning zones with permitted uses and restrictions, terrain, water, and a routed transport graph for walking, cycling, driving and public transport.
    intentresolve against the city modelpropose operationsvalidatepricerequest approvalexecutere-assess

    Planning agent

    Makes the tool usable

    Serves
    the participant
    Does
    turns intent into a rule-checked, priced proposal
    Output
    an operation, awaiting approval
    Runs
    inside the studio

    City agent

    Makes the city agentic

    Serves
    the simulated city
    Does
    lives in the plan: chooses housing, work, travel, consumption, investment
    Output
    behaviour, and a verdict on the city
    Runs
    inside the Economic World Model, after submission

    08 / ASSESSMENT AND SUBMISSION

    Seven perspectives.
    Two files.

    Self-assessment is a working tool, not the final score: it exists so a participant can find the weak part of their own plan before the committee does. Final ranking combines simulated indicators, agent satisfaction and expert review.

    • Housing & city capacityResident capacity, owner/renter split, affordability
    • Mobility & accessibilityCross-boundary, commuting, school and shopping travel times
    • Public-service capacitySchool places, hospital beds, waste collection
    • Safety & resilienceFire and ambulance response, flooding
    • Environment & climateGreen space per capita, emissions, sustainable-mode share
    • Economy & employmentFirm capacity and sector mix, jobs, regional GDP
    • Hong Kong–Shenzhen exchangeEconomic flow, population movement
    .zip

    spatial_plan.zip

    • manifest.jsonformat version, CRS, units, planning boundary, base-map version, plan ID, target year 2035, cost totals
    • plan.gpkgzone functions, buildings and building groups, transport network (new and removed walking, vehicle, bus links)
    • 3d/3D Tiles for visualisation, objects matched to plan.gpkg by shared identifier
    .yaml

    policies.yaml

    policies:
      - id: "P-001"
        name: "Fire service readiness"
        objectives: "Improve emergency access"
        target_type: "fire_station"
        eligibility: ["FS-01", "FS-02", "FS-03"]
        target_count: 3
        support_per_target_hkd: 6000000
        areas: ["Z6-150", "Z6-129", "Z6-232"]
        period: "2030-2035"

    Example abridged from the challenge site. Both files are generated and validated by the system from the participant's final plan. There is no presentation deck in the pipeline.

    The studio organises self-assessment in these seven groups; the nine evaluation dimensions and seven scenario families of the final are on the challenge page: Evaluation ↗

    09 / SCOPE

    What is real,
    and what is staged.

    This is a demonstration of the intended system, and it is worth being precise about which parts are real. The films show the interface and the workflow; they do not expose the implementation. The platform's source code, city data and compiled model are not published.

    Real

    • the V6 city model: 24.3 km², 20,924 existing buildings, zoning, terrain, water, routed network
    • rule checking
    • operation pricing
    • the HK$ 10 bn ledger
    • approval before spend
    • drafts and undo
    • export
    • 2D / 3D synchronisation

    Staged

    • the guided scenarios are fixed, pre-validated scripts with precomputed indicator values
    • model settings is an interface without a live model connection
    • the committee simulation is described rather than executed
    • on this page and in the film, the traffic on the plan is an illustration

    10 / ROADMAP

    From scripts
    to real machinery.

    The demo is not a cut-down version of the final product. It is the interface contract of the final product, with the reasoning replaced by scripts. The roadmap is the order in which the scripts get replaced by real machinery.

    1. Phase 1 · now

      Demo

      Guided scenarios replay fixed, pre-validated steps. The city model, rules, pricing, budget ledger, capacity arithmetic and 2D/3D synchronisation are real; the reasoning is scripted and the indicators are precomputed.

      What it provesA planning decision can be expressed as an intent, priced, approved and reflected in the city within one workspace.

    2. Phase 2

      Live planning loop

      Replace the script with the loop the interface already assumes: a model connection, a stable tool surface, indicators recomputed from the revision, refusal, and exported files that need no hand-repair.

      Exit criterionA participant with no prior knowledge completes a small plan in natural language, with no preset prompts, and the exported files pass validation.

    3. Phase 3

      City agents (EWM)

      The plan becomes a running environment. Household, firm, investor, developer and service agents act inside the submitted spatial and policy conditions, and the assessment stops being a scoreboard of assets and becomes an observation of outcomes.

      Exit criterionThe same submitted plan, run twice, produces the same result; and two plans that score identically on static capacity produce visibly different social outcomes.

    4. Phase 4

      Towards a public platform

      Reproducible scenario packs and reference results, a submission validator anyone can run, a public gallery of strong plans with their simulated outcomes, and reusable region packs.

      Must keepReproducibility.

    Nothing happens to the city without a cost, a rule check and a human decision.

    the discipline every phase has to keep

    THE CHALLENGE · TRACK 3

    Plan the Northern Metropolis
    for 2035.

    Participants act as Chief Planner of a 24.3 km² slice of the Northern Metropolis: the Lok Ma Chau Loop, San Tin and Ngau Tam Mei, with parts of Mai Po and Fairview Park. They deliver a complete 2035 spatial plan plus a policy package inside a fixed HK$ 10,000,000,000 budget. Agentic City is the world behind Track 3 of the Economic World Model Challenge, which opens in December 2026.

    Organised by

    Part of the Agentic City Global Challenge for Hong Kong's Northern Metropolis, organised by Shenzhen Loop Area Institute, The Chinese University of Hong Kong, Shenzhen, and Nanyang Technological University, with the Shenzhen Big Data Research Institute as technical partner.

    Data

    City model compiled from official Hong Kong mapping and water data (iB1000), OpenStreetMap road and transit networks (© OpenStreetMap contributors), and public building and land-use records.

    Read further

    Architecture ↗
    Competition flow ↗
    Roadmap ↗
    Economic World Model paper ↗