Skip to content

feat(llm): cache recent Claude prompt context #910

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
ousamabenyounes wants to merge 1 commit into
base: main
Choose a base branch
from
Open

feat(llm): cache recent Claude prompt context #910

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
20 changes: 16 additions & 4 deletions strix/core/inputs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -25,6 +25,11 @@


DEFAULT_MAX_TURNS = 500
_CACHE_LOCATION_MESSAGE = "message"
_CACHE_LOCATION_TOOL_CONFIG = "tool_config"
_CACHE_ROLE_SYSTEM = "system"
_CACHE_INDEX_PREVIOUS_MESSAGE = -2
_CACHE_INDEX_LAST_MESSAGE = -1


def _accepts_required_tool_choice(model_name: str | None) -> bool:
Expand Down Expand Up @@ -163,7 +168,7 @@ def make_model_settings(
def _prompt_cache_extra_args(model_name: str) -> dict[str, Any] | None:
"""LiteLLM ``cache_control_injection_points`` for Claude prompt caching.

System prompt + rolling last-message breakpoint everywhere; ``tool_config``
System prompt + rolling recent-message breakpoints everywhere; ``tool_config``
only on Bedrock Converse (the only route whose LiteLLM transform consumes
it — elsewhere it leaks onto the wire and native Anthropic 400s). Unmapped
Bedrock models get no points at all: Bedrock rejects the passed-through
Expand All @@ -174,10 +179,17 @@ def _prompt_cache_extra_args(model_name: str) -> dict[str, Any] | None:
if is_bedrock_route(model_name) and not bedrock_route_supports_prompt_caching(model_name):
return None

points: list[dict[str, Any]] = [{"location": "message", "role": "system"}]
points: list[dict[str, Any]] = [
{"location": _CACHE_LOCATION_MESSAGE, "role": _CACHE_ROLE_SYSTEM}
]
if is_bedrock_route(model_name):
points.append({"location": "tool_config"})
points.append({"location": "message", "index": -1})
points.append({"location": _CACHE_LOCATION_TOOL_CONFIG})
points.extend(
[
{"location": _CACHE_LOCATION_MESSAGE, "index": _CACHE_INDEX_PREVIOUS_MESSAGE},
{"location": _CACHE_LOCATION_MESSAGE, "index": _CACHE_INDEX_LAST_MESSAGE},
]
)
return {"cache_control_injection_points": points}


Expand Down
3 changes: 3 additions & 0 deletions tests/test_inputs.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -65,6 +65,7 @@ def test_make_model_settings_enables_prompt_cache_for_bedrock_claude() -> None:
assert _cache_points("bedrock/global.anthropic.claude-opus-4-8") == [
{"location": "message", "role": "system"},
{"location": "tool_config"},
{"location": "message", "index": -2},
{"location": "message", "index": -1},
]

Expand All @@ -80,6 +81,7 @@ def test_make_model_settings_enables_prompt_cache_for_bedrock_claude() -> None:
def test_make_model_settings_enables_prompt_cache_for_non_bedrock_claude(model_name: str) -> None:
assert _cache_points(model_name) == [
{"location": "message", "role": "system"},
{"location": "message", "index": -2},
{"location": "message", "index": -1},
]

Expand Down Expand Up @@ -125,6 +127,7 @@ def test_prompt_cache_kept_for_non_bedrock_claude_even_if_unmapped(monkeypatch:
for model in ("anthropic/claude-brand-new-9", "openrouter/anthropic/claude-brand-new"):
assert _cache_points(model) == [
{"location": "message", "role": "system"},
{"location": "message", "index": -2},
{"location": "message", "index": -1},
]

Expand Down