diff --git a/src/lm_polygraph/model_adapters/whitebox_model_vllm.py b/src/lm_polygraph/model_adapters/whitebox_model_vllm.py
index a3e4e71d8..f95e7fc23 100644
--- a/src/lm_polygraph/model_adapters/whitebox_model_vllm.py
+++ b/src/lm_polygraph/model_adapters/whitebox_model_vllm.py
@@ -4,6 +4,7 @@
 
 import torch
 from typing import List
+from copy import copy
 
 
 class WhiteboxModelvLLM(Model):
@@ -46,8 +47,10 @@ def __init__(
         self.model_type = "vLLMCausalLM"
 
     def generate(self, *args, **kwargs):
-        sampling_params = self.sampling_params
+        sampling_params = copy(self.sampling_params)
         sampling_params.n = kwargs.get("num_return_sequences", 1)
+        if "max_new_tokens" in kwargs:
+            sampling_params.max_tokens = kwargs["max_new_tokens"]
         texts = self.tokenizer.batch_decode(
             kwargs["input_ids"], skip_special_tokens=True
         )