diff --git a/apps/models_provider/impl/minimax_model_provider/credential/llm.py b/apps/models_provider/impl/minimax_model_provider/credential/llm.py
index 807e05edd59..edc92f554e6 100644
--- a/apps/models_provider/impl/minimax_model_provider/credential/llm.py
+++ b/apps/models_provider/impl/minimax_model_provider/credential/llm.py
@@ -1,4 +1,5 @@
 # coding=utf-8
+from copy import deepcopy
 from typing import Dict
 
 from django.utils.translation import gettext_lazy as _, gettext
@@ -86,4 +87,8 @@ def encryption_dict(self, model: Dict[str, object]):
     api_key = forms.PasswordInputField("API Key", required=True)
 
     def get_model_params_setting_form(self, model_name):
-        return MiniMaxLLMModelParams()
+        form = MiniMaxLLMModelParams()
+        if model_name == "MiniMax-M3.1-Flash":
+            form.max_tokens = deepcopy(form.max_tokens)
+            form.max_tokens.attrs["max"] = 524288
+        return form
diff --git a/apps/models_provider/impl/minimax_model_provider/minimax_model_provider.py b/apps/models_provider/impl/minimax_model_provider/minimax_model_provider.py
index 53a8c8f24f6..0ae3327f4ee 100644
--- a/apps/models_provider/impl/minimax_model_provider/minimax_model_provider.py
+++ b/apps/models_provider/impl/minimax_model_provider/minimax_model_provider.py
@@ -36,9 +36,9 @@
 minimax_h3_ttv_model_credential = MiniMaxH3TextToVideoModelCredential()
 minimax_h3_itv_model_credential = MiniMaxH3ImageToVideoModelCredential()
 
-minimax_m2_7 = ModelInfo(
-    "MiniMax-M2.7",
-    _("Latest flagship model with enhanced reasoning and coding. 204K context window"),
+minimax_m3_1_flash = ModelInfo(
+    "MiniMax-M3.1-Flash",
+    _(""),
     ModelTypeConst.LLM,
     minimax_llm_model_credential,
     MiniMaxChatModel,
@@ -51,29 +51,6 @@
     minimax_llm_model_credential,
     MiniMaxChatModel,
 )
-minimax_m2_7_highspeed = ModelInfo(
-    "MiniMax-M2.7-highspeed",
-    _("High-speed version of M2.7 for low-latency scenarios. 204K context window"),
-    ModelTypeConst.LLM,
-    minimax_llm_model_credential,
-    MiniMaxChatModel,
-)
-
-minimax_m2_5 = ModelInfo(
-    "MiniMax-M2.5",
-    _("Peak Performance. Ultimate Value. 204K context window"),
-    ModelTypeConst.LLM,
-    minimax_llm_model_credential,
-    MiniMaxChatModel,
-)
-
-minimax_m2_5_highspeed = ModelInfo(
-    "MiniMax-M2.5-highspeed",
-    _("Same performance, faster and more agile. 204K context window"),
-    ModelTypeConst.LLM,
-    minimax_llm_model_credential,
-    MiniMaxChatModel,
-)
 
 minimax_tts_hd = ModelInfo(
     "speech-2.8-hd",
@@ -107,12 +84,9 @@
 ]
 model_info_manage = (
     ModelInfoManage.builder()
+    .append_model_info(minimax_m3_1_flash)
     .append_model_info(minimax_m3)
-    .append_model_info(minimax_m2_7)
-    .append_model_info(minimax_m2_7_highspeed)
-    .append_model_info(minimax_m2_5)
-    .append_model_info(minimax_m2_5_highspeed)
-    .append_default_model_info(minimax_m2_7)
+    .append_default_model_info(minimax_m3_1_flash)
     .append_model_info(minimax_tts_hd)
     .append_model_info(minimax_tts_turbo)
     .append_default_model_info(minimax_tts_hd)