Platform & Backend Matrix

Check which native and web backends are supported by llamadart and how backend module selection works per platform.

On this page

This page combines platform support and backend-module configuration for llamadart.

The native-assets hook currently pins llamadart-native tag b8157 (hook/build.dart). Module availability below is for that pinned tag.

Platform/architecture coverage#

Platform target	Hook bundle key	`llamadart_native_backends` configurable?	Backend behavior	Status
Android arm64	`android-arm64`	Yes	Defaults: `cpu`, `vulkan` (when present)	Supported
Android x64	`android-x64`	Yes	Defaults: `cpu`, `vulkan` (when present)	Supported
Linux arm64	`linux-arm64`	Yes	Defaults: `cpu`, `vulkan` (when present)	Supported
Linux x64	`linux-x64`	Yes	Defaults: `cpu`, `vulkan` (when present)	Supported
Windows arm64	`windows-arm64`	Yes	Defaults: `cpu`, `vulkan` (when present)	Supported
Windows x64	`windows-x64`	Yes	Defaults: `cpu`, `vulkan` (when present)	Supported
iOS arm64 (device)	`ios-arm64`	No (fixed in hook)	Consolidated runtime: `cpu`, `metal`	Supported
iOS arm64 (simulator)	`ios-arm64-sim`	No (fixed in hook)	Consolidated runtime: `cpu`, `metal`	Supported
iOS x86_64 (simulator)	`ios-x86_64-sim`	No (fixed in hook)	Consolidated runtime: `cpu`, `metal`	Supported
macOS arm64	`macos-arm64`	No (fixed in hook)	Consolidated runtime: `cpu`, `metal`	Supported
macOS x86_64	`macos-x86_64`	No (fixed in hook)	Consolidated runtime: `cpu`, `metal`	Supported
Web (browser)	N/A (JS bridge path)	N/A	Bridge router: `webgpu`, `cpu` fallback	Experimental

Current module availability by bundle (`b8157`)#

Bundle key	Available backend modules in bundle
`android-arm64`	`cpu`, `vulkan`, `opencl`
`android-x64`	`cpu`, `vulkan`, `opencl`
`linux-arm64`	`cpu`, `vulkan`, `blas`
`linux-x64`	`cpu`, `vulkan`, `blas`, `cuda`, `hip`
`windows-arm64`	`cpu`, `vulkan`, `blas`
`windows-x64`	`cpu`, `vulkan`, `blas`, `cuda`
`ios-`, `macos-`	Consolidated Apple runtime ( `cpu` + `metal` path; no split `ggml-*` module selection in hook)

Selector names and aliases#

llamadart_native_backends values are matched against modules discovered in the selected bundle. Current configurable-bundle module names are:

cpu
vulkan
opencl
cuda
blas
hip

Aliases:

vk -> vulkan
ocl -> opencl
open-cl -> opencl

GpuBackend.metal remains valid as a runtime backend preference on Apple targets, but Apple targets are non-configurable in llamadart_native_backends.

Configuring native backend modules#

Use hooks.user_defines.llamadart.llamadart_native_backends:

hooks:
  user_defines:
    llamadart:
      llamadart_native_backends:
        platforms:
          android-arm64:
            backends: [vulkan]
            cpu_profile: full # default; use compact for baseline-only
          linux-x64: [vulkan, cuda]
          windows-x64:
            backends: [vulkan, cuda, blas]

Android arm64 CPU policy keys (platforms.android-arm64):

cpu_profile: full (default): include all Android ARM CPU variant modules.
cpu_profile: compact: include baseline CPU variant module only.
cpu_variants: [...] (advanced): explicit CPU variant list, overrides cpu_profile.

Supported canonical cpu_variants values:

android_armv8.0_1 (baseline)
android_armv8.2_1
android_armv8.2_2
android_armv8.6_1
android_armv9.0_1
android_armv9.2_1
android_armv9.2_2

Variant feature differences:

Variant	Optional feature set used by that module
`android_armv8.0_1`	baseline
`android_armv8.2_1`	`DOTPROD`
`android_armv8.2_2`	`DOTPROD` + `FP16_VECTOR_ARITHMETIC`
`android_armv8.6_1`	`DOTPROD` + `FP16_VECTOR_ARITHMETIC` + `MATMUL_INT8`
`android_armv9.0_1`	`DOTPROD` + `FP16_VECTOR_ARITHMETIC` + `MATMUL_INT8` + `SVE2`
`android_armv9.2_1`	`DOTPROD` + `FP16_VECTOR_ARITHMETIC` + `MATMUL_INT8` + `SVE` + `SME`
`android_armv9.2_2`	`DOTPROD` + `FP16_VECTOR_ARITHMETIC` + `MATMUL_INT8` + `SVE` + `SVE2` + `SME`

Accepted cpu_variants input forms are normalized, for example:

baseline
armv8_6_1
v9_0_1
android-armv9.2_2
libggml-cpu-android_armv8.2_2.so

If cpu_variants contains unknown entries, they are ignored with warnings. If no valid entries remain, selection falls back to cpu_profile (or default full).

Selection and fallback behavior#

Configurable targets start from defaults (cpu, vulkan) if available.
cpu is auto-added as fallback when present in the bundle.
Android arm64 defaults to cpu_profile: full.
cpu_variants (if provided) takes precedence over cpu_profile for Android arm64.
If requested modules are unavailable for a bundle, the hook warns and falls back to defaults.
If defaults are also unavailable, all available modules in that bundle are used as fallback.
Apple targets (ios-*, macos-*) support cpu + metal, but ignore per-backend module config in this hook path because runtime libraries are consolidated.
windows-x64 performs extra runtime dependency validation:
- cuda requires cudart and cublas DLLs.
- blas requires OpenBLAS DLL.
If you change llamadart_native_backends, run flutter clean once to clear stale native-asset outputs.

Edit this page