SeaArt · Background removal
SeaArt Background Remover achieves top-tier hair strand separation with 99.84% foreground recall on the 2048x2048 benchmark anchor, providing ~150 daily free Stamina and pristine 2048px watermark-free transparent PNGs with excellent commercial honesty despite requiring login and taking ~15.5s.
US-West (California, AS2914)
2048 × 2048 · PNG · Open native file ↗
Subsequent pricing and service checks are shown separately without altering historical scores.
One run does not establish all complex scenes, untested paid maxima or long-term uptime.
Plans & historical pricing
SeaArt.ai Stamina and VIP subscription model · 2026-11-01
Base image generation, upscaling, background removal, and community models；Unused stamina resets daily; standard queue; excludes custom LoRA training；Refreshes daily, no renewal；Perpetual access
Fast-track generation, concurrent tasks, and commercial licensing；Pricing visible inside account store (typically approx $9.99/mo)；Auto-renews monthly；Active during billing month
2 plan / billing-period records, all listed; — means that period is not recorded.
Uses daily regenerating Stamina (~150/day free) plus persistent Coins; paid plans include Standard VIP (~$9.99/mo), Pro (~$29.99/mo), and SVIP tiers detailed within account portals.
Verification has gaps; unknown prices and quotas are not inferred.
Cancellation, sources & verification record
Cancellation and refund terms unverified
SeaArt AI Official Homepage and Community ↗ · 2026-11-01 Capture 1 ↗
Pricing recorded on the test date · 2026-11-04
~150 daily free Stamina; Standard VIP ~$9.99/mo
Daily Stamina freemium + VIP plans & coin bundles
Historical context only; not current purchase guidance.
Delivery & access
- Free delivery record
- 2048 × 2048 (100% original size transparent PNG, no watermarks)
- Attached file dimensions
- 2048 × 2048 · PNG
Read directly from the attached file; this does not establish free download availability at that size.
- Watermark
- No watermark
- Free allowance
- ~150 daily Stamina refreshed every 24h (~3.6-4 Stamina per removal, allowing ~35-40 free removals daily)
- Account requirements
- Account required (Google SSO and email registration supported)
- Maximum specification record
- 2048 × 2048 (100% original size transparent PNG, no watermarks)
May include advertised capabilities; this does not establish a tested paid tier.
- Batch & API
- Single-image processing and gallery history management
- Manual refinement
- Not recorded or not verified
Tested result & sample
SeaArt demonstrated top-tier precision on anchor_bgremove_v1, preserving delicate curly hair strands (99.84% recall) with near-zero background residue (0.02% leak) at 2048x2048.
- Edge error MAD
- 0.0941 · Lower is better
- Global error SAD(k)
- 42.688 · Lower is better
- Foreground retention
- 99.840% · Higher is better
- Background residue
- 0.020% · Lower is better
These historical measurements describe alpha error at edges (MAD), total alpha error / 1000 (SAD), foreground retention and background residue. Low-resolution outputs are aligned to the reference under the original protocol; this does not imply a full-resolution export.
- Fixture
- anchor_bgremove_v1
- Recorded environment
- US-West (California, AS2914)
- Network-observed server wait
- 15.49 s
- End-to-end reference
- 15.49 s
Network wait includes round trips, queuing and polling error, not pure inference time. Local timings depend on the device; repetition counts are not standardized in legacy records.
Experience, rules & trust
- Disclosure
- Estimated cost clearly shown before triggering removal (4 Stamina, discounted to 3.6)
- Interaction
- Not recorded or not verified
- Entity record
- STAR CLOUD PTE. LTD.
- Origin
- Not recorded or not verified
- Domain history record
- Not recorded or not verified
- Third-party tracking record
- Google Tag Manager · Cloudflare · Stripe
- External information
- Not recorded or not verified
- Processing
- Cloud GPU async deep learning matting
- Retention statement
- Stored in private gallery with one-click permanent asset deletion
A recorded statement does not verify deletion or all data-processing behavior.
Verifiable audit trail & evidence
Four SHA-256-verified stages link the benchmark fixture, same-run interface screenshots, native exported artifact and archived pricing sources.
anchor_bgremove_v1 · Inspect fixture & file inspector →
Captures entry rules, result view and export disclosure.
Screenshots, native output and official claims support different conclusions. Evidence grades are not product safety certifications.
Evaluation basis & full scores · Overall 83.8/100
Overall = honesty & reputation × 30% + capability × 70%. Component maxima follow this report’s rubric. List scores are rounded for display; original precision is retained here.
Scoring version 2.3.1 · Test fixture anchor_bgremove_v1
Honesty & reputation · 78.5/100 · 30%
| Criterion | Score | Rationale | Basis |
|---|---|---|---|
| Pricing transparency | 25 / 35 | Explanation unavailable | Evidence A → |
| Interaction restraint | 24 / 30 | Explanation unavailable | Evidence A → |
| Background & reputation | 16 / 20 | Explanation unavailable | Evidence A → |
| Engineering & privacy | 13.5 / 15 | Explanation unavailable | Evidence A → |
Capability · 86/100 · 70%
| Criterion | Score | Rationale | Basis |
|---|---|---|---|
| Task quality | 58 / 65 | Transition band error mad_band=0.0941 · Global SAD(k)=42.688 · Foreground recall 99.84% / Background leak 0.020% | Evidence A → |
| Deliverable specifications | 20 / 20 | True Alpha channel (8/8); Max deliverable spec: 2048 × 2048 (100% original size transparent PNG, no watermarks) | Evidence A → |
| Processing speed | 8 / 15 | Network-observed server wait: 15.49s (API mode: async_verified; network upload 0s / end-to-end 15.49s shown for reference only, never penalized) | Evidence A → |
Legacy scores may include advertised maximum capabilities. Read tested output separately from paid capabilities. Records without matching tiers and scope are not ranked together.
Revisions & corrections
- 2026-11-04
Completed real-machine benchmark evaluation and 3-node screenshot verification under RealTest Protocol v2.3.