# QNN LPAI Op Support Revision History

| Introduced in QNN SDK Version | Description |
| --- | --- |
| 2.36.0 | <ul class="simple"><br><li><p>Added support for RmsNorm Op</p></li><br><li><p>Added support Buffer Op Padding Param</p></li><br></ul> |
| 2.35.0 | <ul class="simple"><br><li><p>Added support for StatelessLSTM/GRU Op</p></li><br></ul> |
| 2.34.0 | <ul class="simple"><br><li><p>Added support for BatchToSpace,SpaceToBatch,DepthToSpace,SpaceToDepth,Quantize,Dequantize</p></li><br></ul> |
| 2.32.0 | <ul class="simple"><br><li><p>Added support for Channel_Shuffle op</p></li><br></ul> |
| 2.26.1 | <ul class="simple"><br><li><p>Added Default Support of antialias param for ResizeBilinear Op</p></li><br></ul> |
| 2.24.1 | <ul class="simple"><br><li><p>Add Power op support</p></li><br><li><p>Add LAYERNORM op support</p></li><br><li><p>Add SPLIT op support</p></li><br></ul> |
| 2.23.1 | <ul class="simple"><br><li><dl class="simple"><br><dt>Added support of 8-bit and 16-bit Ops:</dt><dd><ul><br><li><p>Cast, ElementWisePower, ReduceProd, ResizeNearestNeighbor, Split, Layernorm</p></li><br></ul><br></dd><br></dl><br></li><br></ul> |
| 2.16.1 | <ul class="simple"><br><li><dl class="simple"><br><dt>Added Support of 8-bit and 16-bit ElementWiseUnary Op with sub operations:</dt><dd><ul><br><li><p>ElementWiseExp, ElementWiseLog, ElementWiseNeg, ElementWiseSquareRoot</p></li><br></ul><br></dd><br></dl><br></li><br><li><dl class="simple"><br><dt>Added Support of 8-bit and 16-bit ElementWiseBinary Op with sub operations:</dt><dd><ul><br><li><p>ElementWiseAdd, ElementWiseDivide, ElementWiseMultiply, ElementWiseSubtract</p></li><br></ul><br></dd><br></dl><br></li><br><li><dl class="simple"><br><dt>Added Support of 8-bit and 16-bit ElementWiseNeuron Op with sub operations:</dt><dd><ul><br><li><p>Gelu, HardSwish, Relu, ReluMinMax, Sigmoid, Tanh</p></li><br></ul><br></dd><br></dl><br></li><br></ul> |
| 2.13.1 | <ul class="simple"><br><li><p>Added support for Resize and ElementWiseMinimum (8bit/16bit)</p></li><br></ul> |
| 2.12.1 | <ul class="simple"><br><li><p>Added ROI support for Slice and Concat</p></li><br><li><p>Added support for Convert (8bit/16bit)</p></li><br><li><p>Added support for Pool to floor rounding mode</p></li><br></ul> |
| 2.11.1 | <ul class="simple"><br><li><p>Removed Gather op limitation</p></li><br><li><p>Added support deconv with both input and output padding</p></li><br><li><p>Added 4-bit weights support for Conv2d, DepthWiseConv2d, TransposeConv2d</p></li><br><li><p>Added 8-bit support for HardSwish and Gelu</p></li><br><li><p>Added 16-bit support for Argmax, Argmin, Batchnorm, Conv2d, Concat, DepthWiseConv2d,<br>HardSwish, ElementWiseMaximum, ElementWiseAdd, ElementWiseDivide, ElementWiseExp,<br>ElementWiseLog, ElementWiseMultiply, ElementWiseSquareRoot, ElementWiseSubtract,<br>FullyConnected, Gather, Gelu, LogSoftmax, MatMul, Pad, PoolAvg2d, PoolMax2d, Prelu,<br>ReduceMax, ReduceMean, ReduceMin, ReluMinMax,  Reshape, ReduceSum, Relu, Softmax,<br>StridedSlice, Sigmoid, Transpose, Tanh, and TransposeConv2d.</p></li><br></ul> |
| 2.10.1 | <ul class="simple"><br><li><p>Fix for overflow for clip in global AvgPool</p></li><br><li><p>Added 8-bit signed/unsigned support for ReduceMax, ReduceMin, ReduceSum, StridedSlice,<br>Gather, and LogSoftMax.</p></li><br><li><p>Added AvgPool, MaxPool, Pad, TransposeConv2d, and StridedSlice validators.</p></li><br></ul> |

Last Published: Jun 04, 2026

[Previous Topic
QNN HTP Op Definition Revision History](https://docs.qualcomm.com/bundle/publicresource/80-63442-10/topics/htp_opdef_version_history.md) [Next Topic
Common Terminology](https://docs.qualcomm.com/bundle/publicresource/80-63442-10/topics/terminology.md)