// the take
VContainer is a dependency injection container built specifically for Unity, aimed at replacing Zenject/Extenject in performance-sensitive games. It's for Unity developers who want constructor/method injection and scoped lifetimes without the GC pressure and resolve overhead that heavier DI containers bring to a frame budget.
Zero-allocation resolve path matters a lot in Unity where GC spikes show up as visible frame hitches, and the benchmark claims (5-10x faster than Zenject) are backed by a dedicated test workflow in CI. The optional source generator mode avoids reflection entirely at runtime, which is the right tradeoff for a library meant to sit on the hot path. LifetimeScope's async child-scope creation maps cleanly onto Unity's actual loading model (additive scenes, streamed levels) instead of forcing a generic DI lifetime model onto a game engine. The in-editor diagnostics window for inspecting the resolved dependency graph is a real debugging aid that most C# DI containers don't bother shipping.
ECS integration is explicitly labeled beta, so teams building on DOTS can't treat it as a supported path yet. The project has effectively one maintainer (hadashiA), which is a real risk for something this embedded in an app's bootstrapping and object lifetimes - a maintainer gap means PRs and Unity-version compatibility fixes stall. Source generator mode adds a build-time dependency that can be finicky across Unity's Roslyn versions and incremental compilation setup, and debugging codegen failures is harder than debugging a pure-reflection container. There's no built-in migration path or compatibility shim for teams moving off Zenject/Extenject, so adoption on an existing large codebase means a manual rewrite of all bindings.
We read the README, the directory tree and a key source file in November 2026. Nobody pays to be listed. How we judge →