Clarify H3ddle install and model benchmarks

README.md

@@ -21,6 +21,15 @@ for H3ddle's native Apple-Silicon Metal engine. It is published with permission
 from the MiniMax/Hailuo team and remains subject to the included MiniMax H3
 Community License Agreement.
 
+## Use in H3ddle
+
+**FastH3 · VSA is included in H3ddle's managed model catalog.** Users do not
+need to download this repository manually, merge checkpoints, run a conversion
+script, or configure a model folder. In H3ddle, open **Models**, choose
+**FastH3 · VSA** under **Video**, and select **Install**. The app downloads and
+verifies this pinned transformer and reuses shared H3 components that are
+already installed.
+
 The source checkpoint is
 [`FastVideo/FastVideo-FastH3-4-step-Preview-v1-VSA-DataFree`](https://huggingface.co/FastVideo/FastVideo-FastH3-4-step-Preview-v1-VSA-DataFree)
 at revision `b65818d41939b5085451074fe8ca8b799f8d4921`.
@@ -48,15 +57,25 @@ files expected by the app's managed model manifest.
 ## Measured performance
 
 A controlled 512×512, 124-frame comparison on a 32 GB M1 Pro used the same
-prompt, soundscape, seed, four-call schedule, and cache settings:
-
-| Attention | End-to-end time | Time saved | Reduction | Throughput |
-| --- | ---: | ---: | ---: | ---: |
-| Dense | 958.174 s | — | — | 1.000× |
-| Learned VSA | 696.043 s | 262.132 s | 27.4% | 1.377× |
-
-Both outputs were 5.175-second H.264 clips at 24 fps with stereo AAC. These are
-local M1 Pro measurements, not projections for every Apple-Silicon device.
+prompt, soundscape, seed, and cache settings. The measured outputs were
+5.175-second H.264 clips at 24 fps with stereo AAC:
+
+| H3 model | Released passes | Attention | End-to-end time | Compared with VSA | Evidence |
+| --- | ---: | --- | ---: | ---: | --- |
+| Standard | 20 | Dense | ~4,253 s (~70.9 min) | ~6.11× slower | Schedule-normalized estimate |
+| Turbo | 8 | Dense | ~1,782 s (~29.7 min) | ~2.56× slower | Schedule-normalized estimate |
+| FastH3 Dense | 4 | Dense | 958.174 s (16.0 min) | 1.377× slower | Measured conversion baseline |
+| **FastH3 · VSA** | **4** | **Learned VSA, 90% sparse** | **696.043 s (11.6 min)** | **Baseline** | **Measured** |
+
+The Standard and Turbo rows are estimates, not completed matched runs. They
+subtract the 134.439-second fixed pipeline cost observed in a phase-instrumented
+matched-shape receipt, scale only the measured dense transformer work from four
+passes to the released 8- and 20-pass schedules, and then add the fixed cost
+back. The FastH3 Dense and VSA rows are direct end-to-end measurements.
+
+These are local M1 Pro results, not projections for every Apple-Silicon device.
+The direct four-call comparison is a 27.4% end-to-end reduction, or 1.377×
+throughput, for learned VSA versus Dense.
 
 The native deterministic VSA GPU/CPU test measured cosine similarity
 `0.999998633`, mean absolute error `0.000065549`, and maximum error
@@ -68,4 +87,3 @@ FastH3 Preview v1 supports 124–362 frames and requires a short edge of at leas
 480 pixels. It does not support still-image generation, start/end frames,
 ordered image references, or video inpainting. H3ddle enforces these constraints
 in both the app and the native engine.
-