Skip to content
← Models
ClosedReve/10 trustVerified
Reve 2.1 logo

Reve 2.1

The 'images as code' model — layout-first 4K generation with element-level editing, #2 on the Text-to-Image Arena.

Visit official page →

About

Reve 2.1, released July 2026 by the independent lab Reve (replacing Reve 2.0), generates images from structured layouts rather than flat prompts. Before rendering, the model plans a hierarchical scene — every element a node with a position, size, and local description, compared by Reve to HTML or SVG — so each part of the image is addressable and editable like code.

That makes one-element edits possible without re-rolling the whole composition: move a logo, swap a product, fix text, and the rest of the scene stays put. Output is native 4096x4096 (16 megapixels) with strong multilingual in-image text rendering, and the platform wraps generation, layout, effects, and video (via Veo 3.1) in one creative workflow.

It ranks second on the Text-to-Image Arena behind GPT-Image-2. Reve sells subscriptions (Lite $7.99/mo, Pro $19.99/mo) on a creative-energy system with a free tier, plus a separate production API.

Specs

Context window

n/a (image model)

Pricing

subscription: Free tier, Lite $7.99/mo, Pro $19.99/mo (creative-energy system); production API billed separately (provider-published)

Free quota

Free plan with starter creative energy and daily refresh

API access

reve.art platform; production API (separate billing)

Weights

Closed — API only

Released

July 1, 2026

Benchmarks

Independent (Text-to-Image Arena): ranked #2 behind GPT-Image-2. Provider-reported: layout-first architecture claims; no provider benchmark table published.

Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.

Use cases

image-generationimage-editingdesigntypography

Trust score — 7/10

/10 trustVerified

Release facts (July 2026; Reve; replaced 2.0; layout-first 'images as code' approach; native 4K; #2 on Text-to-Image Arena; subscription + separate API) corroborated across The Rundown, AI Radar, Economic Times, and The Planet Tools. Docked for: (1) efficiency and architecture claims lack a published technical report; (2) energy-based pricing is less predictable than per-image billing; (3) proprietary with no weights; free-plan images may be publicly discoverable.

How the trust score is built

Full methodology →
Adoption · stars/downloads, log-scaled vs peers
6.0/10
Maintenance · recency, releases, issue load
9.0/10
Openness · license + docs
2.0/10
Manual review · sanity read + red-flag check
7.0/10
  • · Adoption 6: indie lab; subscription + production API; #2 on Text-to-Image Arena
  • · Maintenance 9: released 2026-07; replaced 2.0; active platform
  • · Openness 2: proprietary, no weights
  • · Manual 7: release facts (Jul 2026; layout-first 'images as code'; native 4K; arena #2) corroborated across The Rundown, AI Radar, Economic Times; docked for no technical report, unpredictable energy pricing, free-plan visibility caveat

Sources