A typed FunctionGraph IR for heterogeneous inference.
flowG is the compiler IR + runtime substrate behind the openIE-dev ecosystem. It takes typed function graphs and dispatches them across CPU, Apple-silicon Metal, WGPU, AMX, and the OpDispatch backend family, with picojoule-budgeted per-op energy measurement built in.
This is the public release surface. Source is private at openIE-dev/flow-g-core: 99 crates spanning the IR, codegen passes, OpDispatch leaves, model runtimes, and benchmarks.
The Graph Coding Era is a 7½-minute narrated visual on a shorter, shared path from intent to silicon. Or read the interactive explainers: how it works · history · the math.
Release binaries + examples + documentation. flowG is free to use software, not an open-source project. See LICENSE for the Business Source License 1.1 terms, which convert to Apache-2.0 four years after each binary's release date.
| Binary | Purpose |
|---|---|
flowg |
The CLI: make-sample, inspect, run (with energy receipts), import PyTorch FX, bench |
flowg-serve |
HTTP serving for compiled graphs (model runtimes, kernels) |
flowg-bench |
Per-op + per-graph benchmarking with energy receipts |
Plus a Rust library, flowg, for embedding.
# via cargo
cargo install flowg
# via cargo-binstall (prebuilt binary)
cargo binstall flowg
# direct download
curl -fsSL https://github.com/openIE-dev/flow-g/releases/latest/download/flowg-$(uname -s)-$(uname -m).tar.gz | tar xzPlatform support:
| Platform | Status |
|---|---|
| macOS arm64 (Apple Silicon) | shipping |
| macOS x86_64 | shipping |
| Linux x86_64 (musl) | shipping |
| Linux aarch64 (musl) | shipping |
| Windows x86_64 | shipping |
# write a sample graph (a two-input `add`)
flowg make-sample add.fg
# inspect the typed graph: its nodes and resource-semantic wires
flowg inspect add.fg
# run it: prints the energy receipt (total µJ, placement, determinism)
flowg run add.fgSee examples/ for runnable programs spanning every backend.
- Typed FunctionGraph IR: not an AST, not a SSA, not a tensor program. Functions are first-class graph nodes with typed signatures, designed for energy-cost-aware dispatch.
- OpDispatch backend: kernels are addressed by
(op, dtype, layout, target)tuples. The runtime picks the lowest-joule kernel that matches. - Heterogeneous by default: CPU + Metal + WGPU + AMX share the same IR. No "GPU port"; just dispatch the same graph at a different target.
- Energy substrate: every op has measured picojoule cost on Apple Silicon (IOReport-backed), NVML for Linux NVIDIA, RAPL for x86_64 Linux. Energy receipts are first-class output.
- 70+ model runtimes: Llama family, Qwen family, Mamba2, Bamba, Jamba, Hunyuan, Gemma 4, LFM2 family, Whisper, Moonshine, V-JEPA, and more, all using the same IR.
flowG is the substrate. The openIE-dev language family compiles to it:
- Lux: general-purpose reactive language → flowG
- JMax: math-native language → flowG
- Joule: energy-aware self-hosted language → flowG
And the substrate is metered by JouleDB, the energy-metered HDC database that records inference receipts.
- Getting started → https://openie-dev.github.io/flow-g
- Examples →
examples/ - API reference → https://docs.rs/flowg
- Source mirror →
openIE-dev/flow-g-core
GitHub Releases: tagged versions with prebuilt binaries for every supported platform, plus a SHA-256 checksums file per release.
- Discussions: Q&A, design discussion
- Issues: bug reports
- Security: see SECURITY.md
- Binaries: Business Source License 1.1; see LICENSE. Free for non-commercial use, internal use by orgs under $1M revenue, security/academic research. Converts to Apache-2.0 four years after each release.
- Documentation: CC-BY-4.0
- Examples: Apache-2.0