Skip to content

fix(model_prices): bedrock Grok 4.6 does not support prompt caching #38699

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

Closed
FahimaGold wants to merge 2 commits into from
Closed

fix(model_prices): bedrock Grok 4.6 does not support prompt caching #38699

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3976584
fix(model_prices): Bedrock Grok 4.6 does not support prompt caching
FahimaGold Aug 28, 2026
fbd65d9
test: move Bedrock Grok caching regression into the mapped hook test
FahimaGold Aug 28, 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
6 changes: 2 additions & 4 deletions litellm/model_prices_and_context_window_backup.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -50346,29 +50346,27 @@
"us.xai.grok-4.6": {
"input_cost_per_token": 2.2e-06,
"output_cost_per_token": 6.6e-06,
"cache_read_input_token_cost": 5.5e-07,
"litellm_provider": "bedrock_converse",
"max_input_tokens": 500000,
"max_output_tokens": 500000,
"max_tokens": 500000,
"mode": "chat",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_prompt_caching": false,
"supports_reasoning": true,
"supports_tool_choice": true,
"supports_vision": true
},
"global.xai.grok-4.6": {
"input_cost_per_token": 2e-06,
"output_cost_per_token": 6e-06,
"cache_read_input_token_cost": 5e-07,
"litellm_provider": "bedrock_converse",
"max_input_tokens": 500000,
"max_output_tokens": 500000,
"max_tokens": 500000,
"mode": "chat",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_prompt_caching": false,
"supports_reasoning": true,
"supports_tool_choice": true,
"supports_vision": true
Expand Down
6 changes: 2 additions & 4 deletions model_prices_and_context_window.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -50346,29 +50346,27 @@
"us.xai.grok-4.6": {
"input_cost_per_token": 2.2e-06,
"output_cost_per_token": 6.6e-06,
"cache_read_input_token_cost": 5.5e-07,
"litellm_provider": "bedrock_converse",
"max_input_tokens": 500000,
"max_output_tokens": 500000,
"max_tokens": 500000,
"mode": "chat",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_prompt_caching": false,
"supports_reasoning": true,
"supports_tool_choice": true,
"supports_vision": true
},
"global.xai.grok-4.6": {
"input_cost_per_token": 2e-06,
"output_cost_per_token": 6e-06,
"cache_read_input_token_cost": 5e-07,
"litellm_provider": "bedrock_converse",
"max_input_tokens": 500000,
"max_output_tokens": 500000,
"max_tokens": 500000,
"mode": "chat",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_prompt_caching": false,
"supports_reasoning": true,
"supports_tool_choice": true,
"supports_vision": true
Expand Down
12 changes: 12 additions & 0 deletions tests/test_litellm/integrations/test_anthropic_cache_control_hook.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1611,6 +1611,18 @@ def test_model_without_caching_support_not_injected(self, monkeypatch):
monkeypatch.setattr(litellm, "enable_anthropic_prompt_caching", True)
assert self._points(model="anthropic.claude-3-5-sonnet-20240620-v1:0", provider="bedrock") == []

@pytest.mark.parametrize("model", ["us.xai.grok-4.6", "global.xai.grok-4.6"])
def test_bedrock_grok_not_injected(self, monkeypatch, local_model_cost_map, model):
"""Bedrock does not support prompt caching for Grok. The cost-map entries must
carry supports_prompt_caching=false so no breakpoints are injected; otherwise
Bedrock rejects the whole request ("You invoked an unsupported model or your
request did not allow prompt caching")."""
from litellm.utils import supports_prompt_caching

monkeypatch.setattr(litellm, "enable_anthropic_prompt_caching", True)
assert supports_prompt_caching(model=model, custom_llm_provider="bedrock") is False
assert self._points(model=model, provider="bedrock") == []

def test_stands_down_when_client_sent_cache_control(self, monkeypatch):
monkeypatch.setattr(litellm, "enable_anthropic_prompt_caching", True)
messages = [
Expand Down
Loading