Catalog

pub const CatalogStatus = enum

Lifecycle status of a catalog entry, controlling visibility and UI treatment.

pub const CatalogEntry = struct

A single managed-model entry describing its identity, download location, hardware requirements, and tested GPU profiles.

pub const FitState = enum

VRAM-fit assessment for a catalog entry against a specific GPU budget.

pub const apple_silicon_profile = "apple-silicon"

Shared GPU profile string used for all Apple Silicon (Metal) devices.

pub const entries = [_]CatalogEntry{

The complete list of ZINC-validated managed models available for download.

pub fn find(id: []const u8) ?*const CatalogEntry

Look up a catalog entry by its short identifier, returning null if not found.

pub fn findForLoadedModel(managed_id: ?[]const u8, model_path: []const u8, display_name: []const u8) ?*const CatalogEntry

Match a loaded model back to a catalog entry, even when it was opened from a raw path instead of a managed-model id.

Resolution order: managed id → parent directory name (for managed-cache `model.gguf` paths) → file-name exact match → file-name compact-case-insensitive match → display-name fuzzy match.

pub fn profileForGpu(config: gpu_detect.GpuConfig) []const u8

Map a detected Vulkan GPU configuration to its catalog profile string.

RDNA4 is split by VRAM tier (≥28 GiB → "amd-rdna4-32gb", ≥14 GiB → "amd-rdna4-16gb", otherwise "amd-rdna4-small"); RDNA3 is split by VRAM tier (≥14 GiB → "amd-rdna3-16gb", otherwise "amd-rdna3-small"); all other vendors map to a single string each.

pub fn profileForMetal() []const u8

Return the catalog profile string for Apple Silicon Metal devices.

pub const nvidia_cuda_profile = "nvidia-cuda"

Shared GPU profile string used for all NVIDIA (CUDA) devices.

pub fn profileForCuda() []const u8

Return the catalog profile string for NVIDIA CUDA devices.

The CUDA backend does not split by VRAM tier the way Vulkan/RDNA4 does — fit is decided by the live `freeMemory()` budget — so a single profile string is sufficient.

pub fn supportsProfile(entry: CatalogEntry, profile: []const u8) bool

Return whether the entry has been tested on the given GPU profile.

pub fn fitsGpu(entry: CatalogEntry, vram_budget_bytes: u64) bool

Return whether the model's VRAM requirement fits within the given budget without enabling MoE offload.

Strict — does not consider the offload escape hatch. Use `fitState` for the offload-aware tri-state assessment.

pub fn requiredVramWithOffload(entry: CatalogEntry) u64

VRAM required when MoE expert tensors are offloaded to host RAM.

Equal to `required_vram_bytes` for dense models (no offloadable tensors).

pub fn fitState(entry: CatalogEntry, vram_budget_bytes: u64) FitState

Tri-state fit assessment that distinguishes "fits as-is" from "fits only with `ZINC_OFFLOAD_MOE_EXPERTS=1`".

Use this to surface the offload escape hatch to users when a model would otherwise look unsupported.

pub fn requiresOffloadToFit(entry: CatalogEntry, vram_budget_bytes: u64) bool

Return true when the model needs `ZINC_OFFLOAD_MOE_EXPERTS=1` to fit (does not fit by itself but does fit with offload enabled).

pub fn supportedOnCurrentGpu(entry: CatalogEntry, profile: []const u8, vram_budget_bytes: u64) bool

Return whether the model is both tested on the given profile and fits in VRAM without MoE offload.

Equivalent to `supportsProfile and fitsGpu`.

pub fn ggufArchForFamily(family: []const u8) ?[]const u8

Map a catalog family string to the GGUF architecture string that models in that family use.

Returns null for unrecognized families — the caller should treat that as an error (a catalog entry with no known architecture mapping). Note that `"qwen3.5"` and `"qwen3.6"` both map to `"qwen35"` because the GGUF file declares the SSM+attention hybrid architecture under that name.