scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-21 00:21:49 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 73

fix(gemini): keep the 2048 cache minimum on Gemini 2.5 Pro only, per Google's live cachedContents API

Browse source
This commit is contained in:
mateo-berri 2026-09-19 00:07:22 -07:00
parent 3289e22834
commit 2303379c20
3 changed files with 6 additions and 9 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
litellm/model_prices_and_context_window_backup.json
View file

@ -25066,7 +25066,6 @@
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"prompt_cache_min_tokens": 2048,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
@ -27139,7 +27138,6 @@
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"prompt_cache_min_tokens": 2048,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,

2
model_prices_and_context_window.json
View file

@ -25066,7 +25066,6 @@
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"prompt_cache_min_tokens": 2048,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,
@ -27139,7 +27138,6 @@
"supports_parallel_function_calling": true,
"supports_pdf_input": true,
"supports_prompt_caching": true,
"prompt_cache_min_tokens": 2048,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_system_messages": true,

11
tests/test_litellm/llms/vertex_ai/context_caching/test_vertex_ai_context_caching.py
View file

@ -1396,14 +1396,15 @@ class TestContextCachingEndpoints:
# Restart the patcher so teardown_method can stop it cleanly
self._token_check_patcher.start()
@pytest.mark.parametrize("model", ["gemini-2.5-flash", "gemini-2.5-pro"])
def test_check_and_create_cache_skips_between_default_and_gemini_2_5_minimum(
self, model, local_model_cost_map
def test_check_and_create_cache_skips_between_default_and_gemini_2_5_pro_minimum(
self, local_model_cost_map
):
"""Gemini 2.5 Flash and Pro need 2048 cached tokens, twice the provider-agnostic default.
"""Gemini 2.5 Pro needs 2048 cached tokens, twice the provider-agnostic default.
Content between the two used to reach Google's cachedContents endpoint and 400.
Content between the two used to reach Google's cachedContents endpoint and 400
with "Cached content is too small".
"""
model = "gemini-2.5-pro"
self._token_check_patcher.stop()
cached_messages = [