Real-time 3D + interactive world models: RTFM + Genie 3 — the property that makes world models creator-researchable

3 sources

World Labs RTFM (real-time feed-forward 3D generation from a single image, no per-scene optimization) and DeepMind Genie 3 GA (interactive world model — user can navigate + interact with the generated environment in real time) are two world model surfaces that ship real-time + interactive milestones in the same week. The two are not equivalent: RTFM is real-time generation (the 3D scene appears fast); Genie 3 is real-time interaction (the user can navigate the scene in real time). Together they make world models creator-researchable: RTFM gives creators a fast 3D asset, Genie 3 gives creators an interactive experience.

Why now

Both shipped in the same week — creators can now research world models as either fast generation (RTFM) or interactive experience (Genie 3).

Why it is worth publishing

Demo potential: side-by-side of a 3D scene where RTFM generates the assets and Genie 3 lets the user navigate them.

Evidence basis

Two independent vendor primary sources + The Decoder weekly roundup

“World Labs RTFM and DeepMind Genie 3 both shipped this week — and together they make world models creator-researchable: RTFM for real-time 3D generation, Genie 3 for real-time interactive navigation.”

Angle

Frame the two as 'real-time generation vs real-time interaction' world model surfaces — RTFM for fast 3D assets, Genie 3 for interactive experience.

Format

Long-form explainer

Demo idea

Record a 10-minute explainer: 3 min intro on 'real-time generation vs interaction' framing, 3 min on RTFM (single-image 3D generation), 3 min on Genie 3 (interactive navigation), 1 min on the bet comparison.

Platform notes

Specific latency targets and per-environment support beyond the captured summary are not extracted; confirm specific surface details against the underlying vendor docs before stating them on the record.

Usable claims

World Labs shipped RTFM — real-time feed-forward 3D world model that generates 3D scenes from a single image without per-scene optimization.
Google DeepMind shipped Genie 3 GA — interactive world model that lets a user navigate and interact with a generated environment in real time.

Evidence pipeline

From the news

Breakdown

RTFM is real-time generation (3D scene appears fast); Genie 3 is real-time interaction (user navigates the scene in real time). This explainer keeps the two real-time properties distinct, rather than treating 'real-time' as a single capability.

Sources

Risks

Use The Decoder and IT之家 as media-type corroboration, but read the underlying vendor docs for any specific capability or benchmark claim before stating it on the record. Verify specific capability claim against the underlying vendor docs and the actual license / pricing matrix before stating it on the record; do not paraphrase per-platform pricing or license terms into specific dollar figures or commercial-use clauses.
Vendor docs confirm real-time / interactive capability but specific latency targets beyond the captured summary are not in this pass. Verify specific capability claim against the underlying vendor docs and the actual license / pricing matrix before stating it on the record; do not paraphrase per-platform pricing or license terms into specific dollar figures or commercial-use clauses.

Demo ideas

Side-by-side: RTFM generates the 3D scene, Genie 3 lets the user navigate it.
Workflow story: 'use RTFM to generate 3D assets + Genie 3 to add interactive navigation', measure time-to-interactive-experience.
Surface matrix: plot RTFM vs Genie 3 on generation speed vs interaction speed vs asset format.