diff --git a/mlir/include/mlir/Dialect/Vector/IR/VectorOps.td b/mlir/include/mlir/Dialect/Vector/IR/VectorOps.td
--- a/mlir/include/mlir/Dialect/Vector/IR/VectorOps.td
+++ b/mlir/include/mlir/Dialect/Vector/IR/VectorOps.td
@@ -110,6 +110,11 @@
     num_batch_dims (see dimension type descriptions below)). For K = 0 (no
     free or batch dimensions), the accumulator and output are a scalar.
 
+    If operands and results are types of different bitwidths, operands are
+    promoted to have the same bitwidth as the result before performing the
+    contraction. For signless integers, the promotion happens via sign
+    extension.
+
     Optional vector mask arguments (produced by CreateMaskOp or ConstantMaskOp)
     specify the dynamic dimension sizes of valid data within the lhs/rhs vector
     arguments.