| What it does | Test and guard AI agents with outcome verification, not reply verification
| On-device moderation that flags hateful, abusive and threatening text in 23 European languages, before the message is sent.
|
|---|
| Stage | Not given
| MVP
|
|---|
| Founded | Not given
| Not given
|
|---|
| Based in | IT
| NL
|
|---|
| Pricing model | Freemium
| Freemium
|
|---|
| Pricing | Free: €0/month with 100 test runs per month; Test: €19/month with 2,000 runs per month; Guard: €99/month with unlimited runs plus live guard features. AI usage is metered separately at cost.
| Free up to 100k monthly active devices with unlimited inference. Custom licenses available.
|
|---|
| Key features | - Isolated sandbox with fresh fake accounts for testing
- 88 scenario templates for refunds, cancellations, pauses and plan changes
- CI gate and CLI with dependency-free deployment
- Policy gate with ALLOW/REVIEW/HOLD/DENY decisions
- Human approval for risky actions
- Independent verification of outcome against ground-truth state
- Guardrails with prompt injection and PII detection
- Per-end-user spending caps in rolling windows
- Append-only evidence log with JSON, CSV and PDF export
- GitHub Action integration for automated testing
| - Three separate scores: hateful, abusive, threat
- Ten target signals under the hate score: race, color, religion, descent, national or ethnic origin, sexual orientation, gender, disability, age, other
- Runs fully on device with no server calls
- Supports 23 European languages from one model file
- 80MB model size optimized for iOS, Android, and browser
- 0.847 macro F1 accuracy on Multilingual HateCheck
- Warn before message is sent
- Comment triage without infrastructure
|
|---|
| What makes it different | - Deterministic verdicts from explicit state checks, not language model decisions
- No-cost deterministic testing; AI features are optional and separately metered
- Integrated sandbox testing and production guarding in one platform
- Idempotent and exactly-once action execution with retry safety
| - Multilingual support across 23 European languages in a single 80MB model
- Complete privacy with no data leaving the device
- Free access up to 100k monthly active devices with unlimited inference
|
|---|
| Community votes | 0 | 0 |
|---|
| DR (Domain Rating by Ahrefs) | Not checked yet ● no change since last check | 37 ● no change since last check |
|---|
| Trust Flow (Majestic) | 0 | 20 |
|---|
| Citation Flow (Majestic) | 0 | 39 |
|---|
| Referring domains (Majestic) | 0 | 616 |
|---|
COLVO
Toxic