Multica vs Symphony
Both turn tickets into agent runs. Multica is its own task board: people assign issues to agents or squads, runs execute on connected machines, and work lands in review. Symphony, from OpenAI, is a spec plus an Elixir reference service that polls an existing tracker, creates a workspace per issue and runs Codex. Multica is source-available; Symphony is Apache-2.0.
Facts side by side
| Fact | Multica | Symphony |
|---|---|---|
| Type | Platform | Orchestrator |
| Languages / SDKs | Go, TypeScript | Elixir |
| License | unknown | Apache-2.0 |
| Pricing model | Open core | Open source, free |
| Orchestration pattern | Supervisor | Other |
| GitHub stars | 51,737 (as of 2026-09-30) | 27,483 (as of 2026-09-30) |
| GitHub forks | 6,718 | 2,849 |
| Last push | 2026-09-30 | 2026-09-15 |
| Latest release | v0.6.0 | v0.0.3 |
| Repository | multica-ai/multica | openai/symphony |
| Website | multica.ai | openai.com |
| Documentation | multica.ai | github.com |
| Last verified | 2026-09-30 | 2026-09-30 |
| MCP support | Partial (checked 2026-09-30) | Unknown (checked 2026-09-30) |
| A2A support | Unknown (checked 2026-09-30) | Unknown (checked 2026-09-30) |
| AG-UI support | Unknown (checked 2026-09-30) | Unknown (checked 2026-09-30) |
| Install verified | Not yet | Not yet |
Choose Multica if
- You want a board that includes people and agents as assignees, with squads whose leader agent delegates to members.
- You use several agent CLIs: Multica's runtime table lists Claude Code, Codex, OpenCode, OpenClaw, Hermes and more.
- You want a self-hosted deployment with Docker Compose or Helm, or a hosted cloud option.
Choose Symphony if
- You already manage work in Linear or GitHub Issues and want a background service driven by ticket states.
- You run Codex: the reference implementation launches Codex in app-server mode only.
- You want an Apache-2.0 spec you can implement yourself, and accept that OpenAI calls it an engineering preview for trusted environments.
Migration notes
The two do not share data. Moving from Symphony to Multica means importing issues into Multica's board and connecting a runtime on each machine; moving the other way means keeping your tracker as the source of truth and writing the workflow prompt and workspace hook Symphony needs. Review differs: Multica has an in-review status on its own board; Symphony's bundled workflow moves a ticket to Human Review and lands it only after a person moves it to Merging. Safety defaults differ too: Multica runs use the daemon user's full permissions and unattended runs answer approval prompts automatically, while Symphony's Elixir implementation rejects Codex approval requests unless configured otherwise. Multica's license forbids offering it as a hosted service without a commercial license.
FAQ
Is Symphony an OpenAI product?
It is an open-source project in the openai GitHub organization. Its README calls it a low-key engineering preview, and the Elixir implementation prototype software for evaluation.
Is Multica open source?
It is source-available. The Multica License is Apache-2.0 plus conditions on hosted or embedded commercial use and branding, so GitHub reports no standard license.
Can Symphony run Claude Code?
Not with the reference implementation, which launches Codex only. The spec is language-agnostic, so another implementation could target a different agent.