+ 
+ 
+ 
+ 
+ 
+
+
+ Quick Start · + What It Does · + HTTP API · + Benchmarks · + OpenRouter Fusion Alternative · + Contributing · + License +
+ **Fusion Engine is a self-hosted Fusion API and OpenRouter Fusion alternative for building reliable AI model ensembles.** Send one prompt to *N* large language models in parallel through [OpenRouter](https://openrouter.ai), collect @@ -19,6 +38,14 @@ chat completions, and benchmarkable fusion results. One prompt in → many models answer → one fused answer out. You control the panel composition, the judge prompts, and where it runs — no vendor lock-in. +| Widget | What Fusion Engine Gives You | +|---|---| +| Fusion API | Self-hosted multi-model fusion with transparent cost, token, and latency data. | +| OpenRouter Fusion | Bring your own OpenRouter models and panel definitions instead of using a black box. | +| Fable Fusion / Fusion Fable | Keyword-aligned multi-model synthesis for agents, research, code, and creative workflows. | +| OpenAI-compatible gateway | `/v1/models` and `/v1/chat/completions` endpoints for existing AI clients. | +| Eval harness | GSM8K, HumanEval, MMLU, GPQA, paired baselines, and cost-aware reporting. | + --- ## What Is Fusion Engine? @@ -430,6 +457,21 @@ templates and keep what moves the metric for *your* workload. --- +## Contributing and Security + +| Resource | Link | +|---|---| +| Contributing guide | [`CONTRIBUTING.md`](CONTRIBUTING.md) | +| Security policy | [`SECURITY.md`](SECURITY.md) | +| CI workflow | [`.github/workflows/ci.yml`](.github/workflows/ci.yml) | +| License | [`LICENSE`](LICENSE) | + +Issues and pull requests should include enough context to reproduce the behavior, +especially for model, panel, judge-template, and benchmark changes. Security +reports should follow the private disclosure path in `SECURITY.md`. + +--- + ## Roadmap - **Web UI** — a browser front-end (on top of the HTTP API) for running fusions