Benchmarks board
This board ranks twelve email platforms by Loopbench overall score for the June 2026 lab window. Higher overall is better. Latency is shown as milliseconds p50 where lower is better and feeds the sending composite defined in methodology. Click any provider for trends and narrative.
| # | Provider | Overall ↓ | Deliverability | Latency p50 | Reliability | Agent | Trend |
|---|---|---|---|---|---|---|---|
| 1 | 9.1 | 8.9 | 300 ms | 9.0 | 9.6 | ||
| 2 | 8.5 | 8.8 | 520 ms | 9.0 | 6.8 | ||
| 3 | 8.4 | 9.0 | 560 ms | 9.2 | 6.5 | ||
| 4 | 8.2 | 8.4 | 390 ms | 8.8 | 7.9 | ||
| 5 | 7.8 | 9.1 | 320 ms | 9.3 | 7.0 | ||
| 6 | 7.6 | 8.5 | 280 ms | 8.9 | 7.4 | ||
| 7 | 7.5 | 8.2 | 740 ms | 8.5 | 6.1 | ||
| 8 | 7.4 | 8.0 | 450 ms | 8.3 | 6.4 | ||
| 9 | 7.2 | 8.1 | 610 ms | 8.2 | 4.8 | ||
| 10 | 7.1 | 7.8 | 700 ms | 8.0 | 5.5 | ||
| 11 | 7.0 | 7.9 | 680 ms | 8.1 | 5.2 | ||
| 12 | 6.9 | 8.0 | 590 ms | 8.0 | 4.9 |
Metric definitions
| Metric | Unit | What we measure |
|---|---|---|
| Deliverability | score | Composite of authentication readiness (SPF, DKIM, DMARC tooling), seed-list inbox placement in our controlled panels, and bounce hygiene signals during the test window. |
| Send latency | ms p50 | Median wall-clock time from authenticated API accept to provider-acknowledged queue for a 1-recipient transactional send in our lab region. |
| Reliability | score | Success rate of scheduled and triggered sends across the window, weighted by recovery after injected soft failures in our harness. |
| Agent surface | score | How completely an external agent can create, schedule, send, and read outcomes via API and/or MCP without a human in the dashboard. |
| Design quality | score | Blind review of on-brand visual fidelity and cross-client rendering for emails produced under a fixed creative brief. |
| Automation depth | score | Ability to express welcome, nurture, and event-triggered flows, including branching and publish reliability from API or UI. |
Category shortlists live on Compare
Sending, design, and automations each get their own ranked HTML tables on /compare. Rank order always follows the displayed score. Brew places in the top two for those three surfaces with earned reasons tied to AI-native generation and agent operation.