scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-10-29 06:33:28 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 267

feat(mistral): add zai-glm-5-3-flash to the cost map (#48270)

Browse source 
Co-authored-by: devin-ai-integration[bot] <158243242+devin-ai-integration[bot]@users.noreply.github.com>
This commit is contained in:
Kerry Lu 2026-10-24 18:20:36 -07:00 • committed by GitHub
parent 2e52e0fb2a
commit de5dc5b238
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
3 changed files with 50 additions and 11 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

23
litellm/model_prices_and_context_window_backup.json
View file

@ -42726,6 +42726,29 @@
"supports_response_schema": true,
"supports_tool_choice": true
},
"mistral/zai-glm-5-3-flash": {
"cache_read_input_token_cost": 3e-08,
"input_cost_per_token": 1.5e-07,
"litellm_provider": "mistral",
"max_input_tokens": 1048576,
"max_output_tokens": 131072,
"max_tokens": 131072,
"mode": "chat",
"output_cost_per_token": 5e-07,
"reasoning_effort_levels": [
"low",
"high",
"max"
],
"source": "https://docs.mistral.ai/models/zai-glm-5-3-flash",
"supports_assistant_prefill": true,
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true
},
"mistral/zai-glm-5": {
"cache_read_input_token_cost": 1.4e-07,
"input_cost_per_token": 1.4e-06,

23
model_prices_and_context_window.json
View file

@ -42726,6 +42726,29 @@
"supports_response_schema": true,
"supports_tool_choice": true
},
"mistral/zai-glm-5-3-flash": {
"cache_read_input_token_cost": 3e-08,
"input_cost_per_token": 1.5e-07,
"litellm_provider": "mistral",
"max_input_tokens": 1048576,
"max_output_tokens": 131072,
"max_tokens": 131072,
"mode": "chat",
"output_cost_per_token": 5e-07,
"reasoning_effort_levels": [
"low",
"high",
"max"
],
"source": "https://docs.mistral.ai/models/zai-glm-5-3-flash",
"supports_assistant_prefill": true,
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true
},
"mistral/zai-glm-5": {
"cache_read_input_token_cost": 1.4e-07,
"input_cost_per_token": 1.4e-06,

15
tests/unit/test_model_prices_and_context_window.py
View file

@ -455,30 +455,23 @@ def is_active_priced_mistral_chat_row(name: str, entry: Mapping[str, object]) ->
)
def cache_read_is_tenth_of_input(entry: Mapping[str, object]) -> bool:
def cache_read_is_below_input(entry: Mapping[str, object]) -> bool:
cache_read: Final = entry.get("cache_read_input_token_cost")
input_cost: Final = entry.get("input_cost_per_token")
return (
isinstance(cache_read, float)
and isinstance(input_cost, (int, float))
and 0 < cache_read < input_cost
and cache_read == pytest.approx(input_cost / 10)
)
return isinstance(cache_read, float) and isinstance(input_cost, (int, float)) and 0 < cache_read < input_cost
@pytest.mark.parametrize("path", (PRICES_PATH, BACKUP_PRICES_PATH), ids=("main", "backup"))
def test_active_mistral_chat_rows_price_cache_reads_below_input(path: Path):
"""A Mistral chat row without a cache-read rate bills cached prompt tokens at zero, so every
active priced row must carry one, and it must be cheaper than a fresh input token. Mistral
bills cached tokens at 10% of the input price for every model (docs.mistral.ai/studio/
conversations/advanced/prompt-caching, read 2026-09-18), so the ratio is checked as well."""
active priced row must carry one, and it must be cheaper than a fresh input token."""
rows: Mapping[str, object] = json.loads(path.read_text())
drifted: Final = [
f"{name}: cache_read={entry.get('cache_read_input_token_cost')} input={entry.get('input_cost_per_token')}"
for name, entry in rows.items()
if isinstance(entry, dict)
and is_active_priced_mistral_chat_row(name, entry)
and not cache_read_is_tenth_of_input(entry)
and not cache_read_is_below_input(entry)
]
assert drifted == []