diff --git a/tests/test_litellm_models.py b/tests/test_litellm_models.py
new file mode 100644
index 00000000..58fa8504
--- /dev/null
+++ b/tests/test_litellm_models.py
@@ -0,0 +1,65 @@
+import importlib.util
+from pathlib import Path
+
+import litellm
+import pytest
+
+
+MODULE_PATH = (
+    Path(__file__).parents[1] / "wdoc" / "utils" / "customs" / "litellm_models.py"
+)
+SPEC = importlib.util.spec_from_file_location("wdoc_litellm_models", MODULE_PATH)
+assert SPEC is not None and SPEC.loader is not None
+litellm_models = importlib.util.module_from_spec(SPEC)
+SPEC.loader.exec_module(litellm_models)
+
+
+def test_registers_minimax_models_with_current_metadata():
+    litellm_models.register_wdoc_models(litellm)
+
+    expected = {
+        "minimax/MiniMax-M3": {
+            "max_tokens": 1_000_000,
+            "input_cost_per_token": 0.6 / 1_000_000,
+            "output_cost_per_token": 2.4 / 1_000_000,
+            "cache_read_input_token_cost": 0.12 / 1_000_000,
+            "cache_creation_input_token_cost": None,
+            "input_modalities": ["text", "image", "video"],
+            "thinking": ["adaptive", "disabled"],
+        },
+        "minimax/MiniMax-M2.7": {
+            "max_tokens": 204_800,
+            "input_cost_per_token": 0.3 / 1_000_000,
+            "output_cost_per_token": 1.2 / 1_000_000,
+            "cache_read_input_token_cost": 0.06 / 1_000_000,
+            "cache_creation_input_token_cost": 0.375 / 1_000_000,
+            "input_modalities": ["text"],
+            "thinking": ["always_on"],
+        },
+    }
+
+    for model_id, metadata in expected.items():
+        assert model_id in litellm.models_by_provider["minimax"]
+        registered = litellm.model_cost[model_id]
+        assert registered["litellm_provider"] == "minimax"
+        assert registered["mode"] == "chat"
+        for key, value in metadata.items():
+            assert registered[key] == value
+
+
+@pytest.mark.parametrize(
+    ("region", "protocol", "expected"),
+    [
+        ("global_en", "openai", "https://api.minimax.io/v1"),
+        ("global_en", "anthropic", "https://api.minimax.io/anthropic"),
+        ("cn_zh", "openai", "https://api.minimaxi.com/v1"),
+        ("cn_zh", "anthropic", "https://api.minimaxi.com/anthropic"),
+    ],
+)
+def test_minimax_endpoint_recipes(region, protocol, expected):
+    assert litellm_models.get_minimax_api_base(region, protocol) == expected
+
+
+def test_minimax_endpoint_recipe_rejects_unknown_selection():
+    with pytest.raises(ValueError, match="Unsupported MiniMax endpoint selection"):
+        litellm_models.get_minimax_api_base("unknown", "openai")
diff --git a/wdoc/utils/customs/litellm_models.py b/wdoc/utils/customs/litellm_models.py
new file mode 100644
index 00000000..008f3cfe
--- /dev/null
+++ b/wdoc/utils/customs/litellm_models.py
@@ -0,0 +1,68 @@
+"""wdoc-specific model metadata registered with LiteLLM."""
+
+from typing import Any, Literal
+
+
+MINIMAX_PROVIDER = "minimax"
+
+MINIMAX_ENDPOINTS = {
+    "global_en": {
+        "openai_base_url": "https://api.minimax.io/v1",
+        "anthropic_base_url": "https://api.minimax.io/anthropic",
+    },
+    "cn_zh": {
+        "openai_base_url": "https://api.minimaxi.com/v1",
+        "anthropic_base_url": "https://api.minimaxi.com/anthropic",
+    },
+}
+
+MINIMAX_MODELS = {
+    "minimax/MiniMax-M3": {
+        "litellm_provider": MINIMAX_PROVIDER,
+        "mode": "chat",
+        "max_tokens": 1_000_000,
+        "max_input_tokens": 1_000_000,
+        "input_cost_per_token": 0.6 / 1_000_000,
+        "output_cost_per_token": 2.4 / 1_000_000,
+        "cache_read_input_token_cost": 0.12 / 1_000_000,
+        "cache_creation_input_token_cost": None,
+        "input_modalities": ["text", "image", "video"],
+        "thinking": ["adaptive", "disabled"],
+        "supports_vision": True,
+        "supports_reasoning": True,
+        "supports_adaptive_thinking": True,
+    },
+    "minimax/MiniMax-M2.7": {
+        "litellm_provider": MINIMAX_PROVIDER,
+        "mode": "chat",
+        "max_tokens": 204_800,
+        "max_input_tokens": 204_800,
+        "input_cost_per_token": 0.3 / 1_000_000,
+        "output_cost_per_token": 1.2 / 1_000_000,
+        "cache_read_input_token_cost": 0.06 / 1_000_000,
+        "cache_creation_input_token_cost": 0.375 / 1_000_000,
+        "input_modalities": ["text"],
+        "thinking": ["always_on"],
+        "supports_reasoning": True,
+    },
+}
+
+
+def get_minimax_api_base(
+    region: Literal["global_en", "cn_zh"],
+    protocol: Literal["openai", "anthropic"] = "openai",
+) -> str:
+    """Return the configured MiniMax base URL for a region and protocol."""
+    try:
+        return MINIMAX_ENDPOINTS[region][f"{protocol}_base_url"]
+    except KeyError as err:
+        raise ValueError(
+            f"Unsupported MiniMax endpoint selection: {region}/{protocol}"
+        ) from err
+
+
+def register_wdoc_models(litellm: Any) -> None:
+    """Register wdoc's model recipes in LiteLLM's cost and provider catalogs."""
+    litellm.register_model(MINIMAX_MODELS)
+    provider_models = litellm.models_by_provider.setdefault(MINIMAX_PROVIDER, set())
+    provider_models.update(MINIMAX_MODELS)
diff --git a/wdoc/wdoc.py b/wdoc/wdoc.py
index 75bd16e2..053b1f2c 100644
--- a/wdoc/wdoc.py
+++ b/wdoc/wdoc.py
@@ -32,6 +32,7 @@
 )
 
 from wdoc.utils.batch_file_loader import batch_load_doc
+from wdoc.utils.customs.litellm_models import register_wdoc_models
 from wdoc.utils.env import env, is_out_piped
 from wdoc.utils.errors import (
     NoDocumentsAfterLLMEvalFiltering,
@@ -120,6 +121,8 @@ def __init__(
         """
         import litellm
 
+        register_wdoc_models(litellm)
+
         if version:
             print(self.VERSION)
             return