Skip to content

fix(gemini): place longest model prefix first in context_windows table (fix #7129) #7146

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Jaymr1031-Z wants to merge 3 commits into
base: main
Choose a base branch
from
Open

fix(gemini): place longest model prefix first in context_windows table (fix #7129) #7146

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
e16531d
fix(gemini): place longest model prefix first in context_windows (fix…
Jaymr1031-Z Aug 28, 2026
80ca7fd
fix(gemini): match specific context windows first
Jaymr1031-Z Aug 30, 2026
7e135b2
Merge branch 'main' of https://github.com/crewAIInc/crewAI into fix/g…
Jaymr1031-Z Aug 30, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
16 changes: 10 additions & 6 deletions lib/crewai/src/crewai/llms/providers/gemini/completion.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1361,24 +1361,28 @@ def get_context_window_size(self) -> int:
f"Context window for {key} must be between {min_context} and {max_context}"
)

# Longest prefix first. Always insert new keys in that order so
# startswith prefers gemini-2.0-flash-thinking over gemini-2.0-flash, etc.
context_windows = {
"gemini-3-pro-preview": 1048576, # 1M tokens
"gemini-2.0-flash": 1048576, # 1M tokens
"gemini-2.0-flash-thinking": 32768,
"gemini-2.0-flash-lite": 1048576,
"gemini-2.0-flash": 1048576, # 1M tokens
"gemini-2.5-flash": 1048576,
"gemini-2.5-pro": 1048576,
"gemini-1.5-pro": 2097152, # 2M tokens
"gemini-1.5-flash": 1048576,
"gemini-1.5-flash-8b": 1048576,
"gemini-1.5-flash": 1048576,
"gemini-1.0-pro": 32768,
"gemma-3-1b": 32000,
"gemma-3-4b": 128000,
"gemma-3-12b": 128000,
"gemma-3-27b": 128000,
"gemma-3-12b": 128000,
"gemma-3-4b": 128000,
"gemma-3-1b": 32000,
}

for model_prefix, size in context_windows.items():
for model_prefix, size in sorted(
context_windows.items(), key=lambda item: len(item[0]), reverse=True
):
if self.model.startswith(model_prefix):
return int(size * CONTEXT_WINDOW_USAGE_RATIO)

Expand Down
12 changes: 12 additions & 0 deletions lib/crewai/tests/test_llm.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -386,6 +386,18 @@ def test_unrecognized_provider_prefix_is_not_stripped() -> None:
)


def test_gemini_thinking_context_window() -> None:
"""Gemini thinking models must not be shadowed by gemini-2.0-flash prefix."""
from crewai.llms.providers.gemini.completion import GeminiCompletion

gemini_thinking = GeminiCompletion(model="gemini-2.0-flash-thinking-exp-01-21")
assert gemini_thinking.get_context_window_size() == int(32768 * CONTEXT_WINDOW_USAGE_RATIO)

gemini_flash = GeminiCompletion(model="gemini-2.0-flash-001")
assert gemini_flash.get_context_window_size() == int(1048576 * CONTEXT_WINDOW_USAGE_RATIO)



@pytest.fixture
def get_weather_tool_schema():
return {
Expand Down