Agent Workflow Platformvsflowise-ai
Two self-hosted AI Agents projects, compared on the numbers selfhost.directory tracks live: score, popularity, release cadence and setup effort.
| Agent Workflow Platform | flowise-ai | |
|---|---|---|
| Score | 39/100 | 74/100 |
| Rank in category | #172 of 209 in AI Agents | #22 of 209 in AI Agents |
| GitHub stars | 105 | 55 561 |
| Stars, last 30 days | +0 | +72 |
| Stable releases, 12 months | 1 | 5 |
| Latest release | v0.1.0 · 23 Aug 2026 | flowise@3.1.4 · 29 Jul 2026 |
| Setup | Advanced | Easy |
| Official Docker image | No | Yes |
| License | MIT | Apache-2.0 |
| Main language | Typescript | Typescript |
Which one should you pick?
flowise-ai scores 74/100 against 39/100 for Agent Workflow Platform, so it is the safer default.
flowise-ai ships more often: 5 stable releases in the past year, against 1 for Agent Workflow Platform.
If you want the quickest setup, flowise-ai is easier to run; Agent Workflow Platform needs more configuration.
flowise-ai is growing faster right now (+72 stars in 30 days).
flowise-ai is archived, so it won't get fixes. Prefer the other one for anything long-lived.
Key differences
- License: Agent Workflow Platform is MIT, flowise-ai is Apache-2.0.
- Only flowise-ai ships an official Docker image.
- Features only Agent Workflow Platform lists: Agent CLI Protocol, Crash Recovery Verification, Cross-platform CI Validation, Docker Compose Deployment and 10 more.
- Features only flowise-ai lists: Broad LLM/vector-store support, Custom MCP tool support, Execution traces and metrics, Horizontal scaling with queue and 6 more.
Agent Workflow Platform details →flowise-ai details →All AI Agents projects →