From 0c7373675f7a4f0ff2a27f1c352ab99e6f931388 Mon Sep 17 00:00:00 2001
From: 9qeklajc <tthefux@gmail.com>
Date: Sat, 27 Jun 2026 22:53:35 +0200
Subject: [PATCH] solidify logic

---
 routstr/payment/models.py        |  8 +++++-
 tests/unit/test_cache_pricing.py | 42 ++++++++++++++++++++++++++++++++
 2 files changed, 49 insertions(+), 1 deletion(-)

diff --git a/routstr/payment/models.py b/routstr/payment/models.py
index db477b93..e6ea643d 100644
--- a/routstr/payment/models.py
+++ b/routstr/payment/models.py
@@ -243,7 +243,13 @@ def _row_to_model(
     # without this, ``_row_to_model`` bills cache reads at the full input rate —
     # the ``_apply_provider_fee_to_model`` path backfills, but the override path
     # used for admin-configured providers did not.
-    parsed_pricing = backfill_cache_pricing(row.id, parsed_pricing)
+    #
+    # Key on ``forwarded_model_id`` (the actual upstream model name litellm
+    # prices) when set: an alias row (id="local-alias",
+    # forwarded_model_id="deepseek-v4-flash") would otherwise look up the alias
+    # and miss the cache rate.
+    pricing_model_id = getattr(row, "forwarded_model_id", None) or row.id
+    parsed_pricing = backfill_cache_pricing(pricing_model_id, parsed_pricing)
 
     if apply_provider_fee:
         parsed_pricing = Pricing.parse_obj(
diff --git a/tests/unit/test_cache_pricing.py b/tests/unit/test_cache_pricing.py
index 14bb884c..cfb7e9a6 100644
--- a/tests/unit/test_cache_pricing.py
+++ b/tests/unit/test_cache_pricing.py
@@ -196,6 +196,48 @@ def test_row_to_model_backfills_cache_rate() -> None:
     assert model.sats_pricing.input_cache_read > 0
 
 
+def test_row_to_model_backfills_via_forwarded_model_id() -> None:
+    """An alias row (id != forwarded_model_id) must backfill cache rates from
+    the *forwarded* model name — the real upstream model litellm prices —
+    not the alias id, which litellm doesn't know."""
+    import json
+
+    from routstr.core.db import ModelRow
+    from routstr.payment.models import _row_to_model
+
+    row = ModelRow(
+        id="local-alias",  # litellm has no such key
+        name="local-alias",
+        created=0,
+        description="",
+        context_length=1000000,
+        architecture=json.dumps(
+            {
+                "modality": "text",
+                "input_modalities": ["text"],
+                "output_modalities": ["text"],
+                "tokenizer": "unknown",
+                "instruct_type": None,
+            }
+        ),
+        pricing=json.dumps({"prompt": 1.4e-07, "completion": 2.8e-07}),
+        enabled=True,
+        upstream_provider_id=1,
+        forwarded_model_id="deepseek-v4-flash",
+    )
+
+    with patch(
+        "routstr.payment.models.sats_usd_price", return_value=5.0e-5
+    ):
+        model = _row_to_model(row, apply_provider_fee=True, provider_fee=1.0)
+
+    litellm_read = litellm.model_cost["deepseek-v4-flash"][
+        "cache_read_input_token_cost"
+    ]
+    assert model.pricing.input_cache_read == pytest.approx(litellm_read)
+    assert model.pricing.input_cache_read < model.pricing.prompt
+
+
 # ============================================================================
 # calculate_cost — cached tokens billed at cache rates
 # ============================================================================