The AI arena is free today

Open Superagent
Back to blog
Model Release·Image·Multimodal·API

Grok Imagine Image 2.0: xAI's Precision Image Model

Grok Imagine Image 2.0 (grok-imagine-image-2.0): xAI Quality Mode image model—text+image in, image out, $0.04/image, region edit, up to 5 refs. GA Aug 7, 2026 on grok.com/imagine and API.

Sebastian Crossa
Sebastian Crossa
Co-Founder @ LLM Stats
·2 min read
Grok Imagine Image 2.0: xAI's Precision Image Model

At a Glance

  • ID: grok-imagine-image-2.0
  • Org: xAI (SpaceXAI)
  • Modalities: Text, image in; image out
  • Consumer: Quality Mode on grok.com/imagine + mobile apps (GA Aug 7, 2026)
  • API price: $0.04 / output image
  • Rate limit: 5 requests/sec
  • Regions: us-east-1, us-west-2
  • Peers: grok-imagine-image ($0.02), grok-imagine-image-quality ($0.05), grok-imagine-image-pro

Pricing (docs)

ModelCost
grok-imagine-image-2.0$0.04 / image
grok-imagine-image$0.02 / image
grok-imagine-image-quality$0.05 / image

Source: https://docs.x.ai/developers/pricing and model page.

What's New

  1. Precise editing: Magic wand region edits, segmentation, background removal (transparent export), multi-ref editing (up to 5 input images), smart resize across common aspect ratios (1:2 through 2:1).
  2. Design / typography focus: Vendor pitch on dense multi-part layouts and sharp small text.
  3. Templates: Ready workflows for photo edit, product shots, headshots, icons, game assets, merch, etc.
  4. World-building for video: Character / location / prop sets that hold style across stills (consumer story).

Vendor ranking claims (labeled)

Launch post cites Arena Image Edit and Text-to-Image leaderboards (as of Aug 7, 2026): Image 2.0 ranks second overall Elo on both. xAI models listed under SpaceXAI. Vendor-cited leaderboard snapshot, not LLM Stats verified. Do not treat as a permanent rank.

When to Use It

Good fit: Iterative creative work (edit loops, product / marketing stills, typography-heavy posters), API jobs that need the Quality Mode stack at $0.04/image.

Cost note: Twice the per-image price of base grok-imagine-image. Prefer the cheaper ID for drafts; step up to 2.0 for fidelity / edit workflows.

Caveats

  • Consumer "Quality Mode" naming vs API slug grok-imagine-image-2.0 (and separate grok-imagine-image-quality price row) can confuse catalogs. Keep IDs distinct.
  • Arena Elo claims are time-stamped snapshots; re-check before promoting ranks.
  • Rate limit is tight (5 rps) vs video's 10 rps.

Sources

Questions

Frequently Asked Questions

  • August 7, 2026 (GA on grok.com/imagine and mobile apps).

  • grok-imagine-image-2.0. On the consumer product it is Quality Mode; keep that distinct from the separate price row grok-imagine-image-quality.
  • $0.04 per output image. Base grok-imagine-image is $0.02; grok-imagine-image-quality is $0.05. Source: docs.x.ai/developers/pricing.
  • Text and image in; image out. Editing features include region edit, segmentation, background removal, up to 5 reference images, and smart resize (1:2 through 2:1).
  • 5 requests/sec. Regions: us-east-1, us-west-2.
  • No. Launch post cites Arena Image Edit and Text-to-Image leaderboards as of Aug 7, 2026 (second overall Elo on both). Treat as a vendor-cited snapshot, not LLM Stats verified.

Continue Reading