BIRME · Background removal
BIRME employs an innovative client-side WebGPU/WASM matting engine. Processing the 2048×2048 curly hair portrait took 28.5s entirely within the browser, achieving 99.9% foreground recall and only 0.8% background residue. Completely free without login, it exports full-resolution unwatermarked transparent PNGs with total privacy.
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
BIRME 100% free client-side batch image processing and compression · 2026-11-01
Smart crop, format conversion (JPG/PNG/WebP/AVIF), quality compression, watermark, AI background removal and upscaling；Pure client-side execution; very large batches may consume significant browser memory；Permanently free, no renewal；Perpetual access
1 plan / billing-period records, all listed; — means that period is not recorded.
All batch resizing, compression, AI background removal, and upscaling are 100% free with voluntary donation support and zero paid plans.
Cancellation, sources & verification record
Cancellation and refund terms unverified
BIRME Official Studio ↗ · 2026-11-01 Capture 1 ↗
Pricing recorded on the test date · 2026-11-04
100% free with no paid tiers or quota limits
free
Historical context only; not current purchase guidance.
Delivery & access
- Free delivery record
- 2048x2048 (PNG full resolution transparent deliverable)
- Attached file dimensions
- 2048 × 2048 · PNG
Read directly from the attached file; this does not establish free download availability at that size.
- Watermark
- Not recorded or not verified
- Free allowance
- Completely free without login; unlimited batch image processing constrained only by client memory
- Account requirements
- Direct usage without login
- Maximum specification record
- 2048x2048 (PNG full resolution transparent deliverable)
May include advertised capabilities; this does not establish a tested paid tier.
- Batch & API
- Not recorded or not verified
- Manual refinement
- Not recorded or not verified
Tested result & sample
BIRME in-browser AI matting ran locally in 28.5s, accurately separating subject from bookshelf and exporting a 5.78 MB transparent PNG.
- Edge error MAD
- 0.1359 · Lower is better
- Global error SAD(k)
- 56.05 · Lower is better
- Foreground retention
- 99.900% · Higher is better
- Background residue
- 0.804% · 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
- 28.5 s
- End-to-end reference
- 28.5 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
- Not recorded or not verified
- Interaction
- Not recorded or not verified
- Entity record
- BIRME Project
- Origin
- United Kingdom
- Domain history record
- Established in 2013
- Third-party tracking record
- Google Analytics
- External information
- Continuously operating for over 13 years since 2013, with over 50,000 monthly active users and widely praised for client-side privacy.
- Processing
- Not recorded or not verified
- Retention statement
- Zero files retained on server; memory purged upon closing browser tab
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 83.1/100
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 · 96/100 · 30%
| Criterion | Score | Rationale | Basis |
|---|---|---|---|
| Pricing transparency | 35 / 35 | Explanation unavailable | Evidence A → |
| Interaction restraint | 28 / 30 | Explanation unavailable | Evidence A → |
| Background & reputation | 18 / 20 | Explanation unavailable | Evidence A → |
| Engineering & privacy | 15 / 15 | Explanation unavailable | Evidence A → |
Capability · 77.5/100 · 70%
| Criterion | Score | Rationale | Basis |
|---|---|---|---|
| Task quality | 46.5 / 65 | Transition band error mad_band=0.1359 · Global SAD(k)=56.05 · Foreground recall 99.90% / Background leak 0.804% | Evidence A → |
| Deliverable specifications | 20 / 20 | True Alpha channel (8/8); Max deliverable spec: 2048x2048 (PNG full resolution transparent deliverable) | Evidence A → |
| Processing speed | 11 / 15 | Network-observed server wait: 28.5s (API mode: client_side; network upload 0s / end-to-end 28.5s 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.