diff --git a/vllm/model_executor/models/minimax_m2.py b/vllm/model_executor/models/minimax_m2.py
index 3e6a9add9ec4..88705a9cde2c 100644
--- a/vllm/model_executor/models/minimax_m2.py
+++ b/vllm/model_executor/models/minimax_m2.py
@@ -145,6 +145,7 @@ def forward(self, hidden_states: torch.Tensor) -> torch.Tensor:
 class MiniMaxM2Attention(nn.Module):
     def __init__(
         self,
+        config: PretrainedConfig,
         hidden_size: int,
         num_heads: int,
         num_kv_heads: int,
@@ -199,9 +200,14 @@ def __init__(
             prefix=f"{prefix}.o_proj",
         )
 
+        final_rotary_dim = rotary_dim
+        # https://github.com/vllm-project/vllm/pull/30384
+        if hasattr(config, "partial_rotary_factor"):
+            final_rotary_dim = self.head_dim
+
         self.rotary_emb = get_rope(
             self.head_dim,
-            rotary_dim=self.head_dim,
+            rotary_dim=final_rotary_dim,
             max_position=max_position_embeddings,
             rope_parameters=rope_parameters,
         )
@@ -260,6 +266,7 @@ def __init__(
 
         self.layer_idx = layer_idx
         self.self_attn = MiniMaxM2Attention(
+            config=config,
             hidden_size=self.hidden_size,
             num_heads=config.num_attention_heads,
             num_kv_heads=config.num_key_value_heads,