From 569739912520f96567abf236c71cf57aeaacf260 Mon Sep 17 00:00:00 2001
From: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
Date: Wed, 28 Oct 2026 12:40:28 +0000
Subject: [PATCH] revert(cost-map): leave mistral-large-4 rows to #48825

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
---
 litellm/model_prices_and_context_window_backup.json | 12 ++++++------
 model_prices_and_context_window.json                | 12 ++++++------
 tests/unit/test_mistral_large_4_model_metadata.py   | 12 +++++-------
 3 files changed, 17 insertions(+), 19 deletions(-)

diff --git a/litellm/model_prices_and_context_window_backup.json b/litellm/model_prices_and_context_window_backup.json
index c51769920b5..4e6b4d51712 100644
--- a/litellm/model_prices_and_context_window_backup.json
+++ b/litellm/model_prices_and_context_window_backup.json
@@ -42055,13 +42055,13 @@
         "mode": "embedding"
     },
     "mistral/mistral-large-4": {
-        "cache_read_input_token_cost": 1.4e-07,
-        "input_cost_per_token": 1.36e-06,
+        "cache_read_input_token_cost": 6.8e-08,
+        "input_cost_per_token": 6.8e-07,
         "litellm_provider": "mistral",
         "max_input_tokens": 1048576,
         "max_output_tokens": 1048576,
         "mode": "chat",
-        "output_cost_per_token": 4.18e-06,
+        "output_cost_per_token": 2.09e-06,
         "reasoning_effort_levels": [
             "none",
             "high"
@@ -42076,13 +42076,13 @@
         "supports_vision": true
     },
     "mistral/mistral-large-4-0": {
-        "cache_read_input_token_cost": 1.4e-07,
-        "input_cost_per_token": 1.36e-06,
+        "cache_read_input_token_cost": 6.8e-08,
+        "input_cost_per_token": 6.8e-07,
         "litellm_provider": "mistral",
         "max_input_tokens": 1048576,
         "max_output_tokens": 1048576,
         "mode": "chat",
-        "output_cost_per_token": 4.18e-06,
+        "output_cost_per_token": 2.09e-06,
         "reasoning_effort_levels": [
             "none",
             "high"
diff --git a/model_prices_and_context_window.json b/model_prices_and_context_window.json
index c51769920b5..4e6b4d51712 100644
--- a/model_prices_and_context_window.json
+++ b/model_prices_and_context_window.json
@@ -42055,13 +42055,13 @@
         "mode": "embedding"
     },
     "mistral/mistral-large-4": {
-        "cache_read_input_token_cost": 1.4e-07,
-        "input_cost_per_token": 1.36e-06,
+        "cache_read_input_token_cost": 6.8e-08,
+        "input_cost_per_token": 6.8e-07,
         "litellm_provider": "mistral",
         "max_input_tokens": 1048576,
         "max_output_tokens": 1048576,
         "mode": "chat",
-        "output_cost_per_token": 4.18e-06,
+        "output_cost_per_token": 2.09e-06,
         "reasoning_effort_levels": [
             "none",
             "high"
@@ -42076,13 +42076,13 @@
         "supports_vision": true
     },
     "mistral/mistral-large-4-0": {
-        "cache_read_input_token_cost": 1.4e-07,
-        "input_cost_per_token": 1.36e-06,
+        "cache_read_input_token_cost": 6.8e-08,
+        "input_cost_per_token": 6.8e-07,
         "litellm_provider": "mistral",
         "max_input_tokens": 1048576,
         "max_output_tokens": 1048576,
         "mode": "chat",
-        "output_cost_per_token": 4.18e-06,
+        "output_cost_per_token": 2.09e-06,
         "reasoning_effort_levels": [
             "none",
             "high"
diff --git a/tests/unit/test_mistral_large_4_model_metadata.py b/tests/unit/test_mistral_large_4_model_metadata.py
index a945a5fdf93..f45c53920dd 100644
--- a/tests/unit/test_mistral_large_4_model_metadata.py
+++ b/tests/unit/test_mistral_large_4_model_metadata.py
@@ -32,7 +32,9 @@ def test_metadata_and_backup(model):
     assert info["litellm_provider"] == "mistral"
     assert info["mode"] == "chat"
     assert info["max_input_tokens"] == 1048576
-    assert 0 < info["cache_read_input_token_cost"] < info["input_cost_per_token"] < info["output_cost_per_token"]
+    assert info["input_cost_per_token"] == 6.8e-07
+    assert info["output_cost_per_token"] == 2.09e-06
+    assert info["cache_read_input_token_cost"] == 6.8e-08
     assert info["reasoning_effort_levels"] == ["none", "high"]
     for capability in (
         "supports_vision",
@@ -57,16 +59,12 @@ def test_reasoning_effort(model, effort):
 
 @pytest.mark.parametrize("model", MODELS)
 def test_completion_cost(model):
-    info = litellm.get_model_info(model)
-    input_rate = info["input_cost_per_token"]
-    output_rate = info["output_cost_per_token"]
-    cache_read_rate = info["cache_read_input_token_cost"]
     uncached_response = litellm.ModelResponse(
         model=model.split("/", 1)[1],
         usage={"prompt_tokens": 1000, "completion_tokens": 100, "total_tokens": 1100},
     )
     assert litellm.completion_cost(completion_response=uncached_response, model=model) == pytest.approx(
-        1000 * input_rate + 100 * output_rate
+        1000 * 6.8e-07 + 100 * 2.09e-06
     )
     response = litellm.ModelResponse(
         model=model.split("/", 1)[1],
@@ -78,7 +76,7 @@ def test_completion_cost(model):
         },
     )
     assert litellm.completion_cost(completion_response=response, model=model) == pytest.approx(
-        400 * input_rate + 600 * cache_read_rate + 100 * output_rate
+        400 * 6.8e-07 + 600 * 6.8e-08 + 100 * 2.09e-06
     )