test(e2e): derive cache rates from first principles and ungate all_components cases

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

tests/e2e/cost_calculation/cases.json

@@ -181,9 +181,6 @@
         "audio_output_tokens": 3
       },
       "requires_rates": [
-        "cache_read_input_token_cost",
-        "cache_creation_input_token_cost",
-        "cache_creation_input_token_cost_above_1hr",
         "output_cost_per_reasoning_token",
         "input_cost_per_audio_token",
         "output_cost_per_audio_token"
@@ -193,7 +190,6 @@
     {
       "name": "all_components_fireworks",
       "usage": {"fresh_input_tokens": 80, "cache_read_tokens": 40, "output_tokens": 25},
-      "requires_rates": ["cache_read_input_token_cost"],
       "wires": ["fireworks_chat"]
     },
     {
@@ -205,11 +201,6 @@
         "cache_write_1h_tokens": 10,
         "output_tokens": 25
       },
-      "requires_rates": [
-        "cache_read_input_token_cost",
-        "cache_creation_input_token_cost",
-        "cache_creation_input_token_cost_above_1hr"
-      ],
       "wires": ["anthropic_messages", "bedrock_converse"]
     },
     {
@@ -222,11 +213,6 @@
         "output_tokens": 25
       },
       "stream": true,
-      "requires_rates": [
-        "cache_read_input_token_cost",
-        "cache_creation_input_token_cost",
-        "cache_creation_input_token_cost_above_1hr"
-      ],
       "wires": ["anthropic_messages"]
     },
     {
@@ -240,7 +226,6 @@
         "audio_output_tokens": 3
       },
       "requires_rates": [
-        "cache_read_input_token_cost",
         "output_cost_per_reasoning_token",
         "input_cost_per_audio_token",
         "output_cost_per_audio_token"
@@ -250,7 +235,7 @@
     {
       "name": "all_components_responses",
       "usage": {"fresh_input_tokens": 80, "cache_read_tokens": 40, "output_tokens": 25, "reasoning_tokens": 15},
-      "requires_rates": ["cache_read_input_token_cost", "output_cost_per_reasoning_token"],
+      "requires_rates": ["output_cost_per_reasoning_token"],
       "wires": ["openai_responses"]
     }
   ]

tests/e2e/cost_calculation/expected.json

@@ -1686,6 +1686,13 @@
     "prompt_tokens": 100,
     "spend": 0.0216
   },
+  "meta.llama4-maverick-17b-instruct-v1:0|all_components_anthropic": {
+    "completion_tokens": 25,
+    "input_cost": 0.0285,
+    "output_cost": 0.0095,
+    "prompt_tokens": 150,
+    "spend": 0.038
+  },
   "meta.llama4-maverick-17b-instruct-v1:0|basic": {
     "completion_tokens": 40,
     "input_cost": 0.0228,

tests/e2e/cost_calculation/generate_expected.py

@@ -94,11 +94,22 @@ def expected_breakdown(model: FrontierModel, case: Case) -> ExpectedCost:
         or rates.output_cost_per_token
         or 0.0
     )
+    write_rate: Final = (
+        rates.cache_creation_input_token_cost
+        if rates.cache_creation_input_token_cost is not None
+        else in_rate
+    )
     input_cost: Final = (
         u.fresh_input_tokens * in_rate
-        + u.cache_read_tokens * (rates.cache_read_input_token_cost or 0.0)
-        + u.cache_write_5m_tokens * (rates.cache_creation_input_token_cost or 0.0)
-        + u.cache_write_1h_tokens * (rates.cache_creation_input_token_cost_above_1hr or 0.0)
+        + u.cache_read_tokens
+        * (rates.cache_read_input_token_cost if rates.cache_read_input_token_cost is not None else in_rate)
+        + u.cache_write_5m_tokens * write_rate
+        + u.cache_write_1h_tokens
+        * (
+            rates.cache_creation_input_token_cost_above_1hr
+            if rates.cache_creation_input_token_cost_above_1hr is not None
+            else write_rate
+        )
         + u.audio_input_tokens * (rates.input_cost_per_audio_token or 0.0)
     )
     output_cost: Final = (