scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-11 22:51:28 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 123

test(registry): cover nemotron reasoning, v4-flash vision and xai/groq deprecation dates

Browse source 
Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>
This commit is contained in:
mateo 2026-09-11 13:28:35 +00:00
parent 4382b86b0f
commit 8762c664b7
1 changed files with 44 additions and 12 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

56
tests/test_litellm/test_utils.py
View file

@ -198,6 +198,8 @@ def test_get_model_info_resolves_provider_prefixed_model_ids(local_model_cost_ma
("perplexity/perplexity/kimi-k3", True),
("perplexity/perplexity/deepseek-v4-flash-0731", True),
("perplexity/perplexity/kimi-k2.7-code", False),
("perplexity/perplexity/nemotron-3.5-lightning-30b-a3b", True),
("perplexity/perplexity/nemotron-3-ultra-550b-a55b", True),
):
assert litellm.supports_reasoning(model=model) is reasoning, model
@ -209,6 +211,18 @@ def test_get_model_info_resolves_provider_prefixed_model_ids(local_model_cost_ma
assert via_provider["output_cost_per_token"] == 4.4e-06
assert via_provider["mode"] == "responses"
lightning = litellm.get_model_info(
model="perplexity/nemotron-3.5-lightning-30b-a3b", custom_llm_provider="perplexity"
)
assert lightning["key"] == "perplexity/perplexity/nemotron-3.5-lightning-30b-a3b"
assert lightning["input_cost_per_token"] == 1.15e-08
assert lightning["output_cost_per_token"] == 1.7e-07
assert lightning["cache_read_input_token_cost"] == 1.15e-09
assert lightning["mode"] == "responses"
ultra = litellm.get_model_info(model="perplexity/perplexity/nemotron-3-ultra-550b-a55b")
assert ultra["key"] == "perplexity/perplexity/nemotron-3-ultra-550b-a55b"
def test_get_model_info_strips_openai_finetune_ids_without_a_custom_suffix(local_model_cost_map):
info = litellm.get_model_info(model="ft:gpt-4o-2024-08-06:my-org::abc123", custom_llm_provider="openai")
@ -4088,9 +4102,9 @@ def test_deepseek_v4_models_in_cost_map():
model_cost = json.load(f)
# --- bare model names ---
for key, expected_input, expected_output, expected_cache in [
("deepseek-v4-flash", 3e-07, 1.2e-06, 6e-09),
("deepseek-v4-pro", 1.32e-06, 3.96e-06, 4.4e-08),
for key, expected_input, expected_output, expected_cache, expected_vision in [
("deepseek-v4-flash", 3e-07, 1.2e-06, 6e-09, True),
("deepseek-v4-pro", 1.32e-06, 3.96e-06, 4.4e-08, False),
]:
info = model_cost.get(key)
assert info is not None, f"{key} missing from model_prices_and_context_window.json"
@ -4102,11 +4116,12 @@ def test_deepseek_v4_models_in_cost_map():
assert info["max_input_tokens"] == 1_000_000
assert info["supports_function_calling"] is True
assert info["supports_tool_choice"] is True
assert info.get("supports_vision", False) is expected_vision
# --- provider-prefixed names ---
for key, expected_input, expected_output, expected_cache in [
("deepseek/deepseek-v4-flash", 3e-07, 1.2e-06, 6e-09),
("deepseek/deepseek-v4-pro", 1.32e-06, 3.96e-06, 4.4e-08),
for key, expected_input, expected_output, expected_cache, expected_vision in [
("deepseek/deepseek-v4-flash", 3e-07, 1.2e-06, 6e-09, True),
("deepseek/deepseek-v4-pro", 1.32e-06, 3.96e-06, 4.4e-08, False),
]:
info = model_cost.get(key)
assert info is not None, f"{key} missing from model_prices_and_context_window.json"
@ -4117,6 +4132,7 @@ def test_deepseek_v4_models_in_cost_map():
assert info["cache_read_input_token_cost"] == expected_cache
assert info["supports_function_calling"] is True
assert info["supports_tool_choice"] is True
assert info.get("supports_vision", False) is expected_vision
def test_deepseek_v4_models_in_backup_cost_map():
@ -4132,9 +4148,9 @@ def test_deepseek_v4_models_in_backup_cost_map():
model_cost = json.load(f)
# --- bare model names ---
for key, expected_input, expected_output, expected_cache in [
("deepseek-v4-flash", 3e-07, 1.2e-06, 6e-09),
("deepseek-v4-pro", 1.32e-06, 3.96e-06, 4.4e-08),
for key, expected_input, expected_output, expected_cache, expected_vision in [
("deepseek-v4-flash", 3e-07, 1.2e-06, 6e-09, True),
("deepseek-v4-pro", 1.32e-06, 3.96e-06, 4.4e-08, False),
]:
info = model_cost.get(key)
assert info is not None, f"{key} missing from backup JSON"
@ -4144,11 +4160,12 @@ def test_deepseek_v4_models_in_backup_cost_map():
assert info["output_cost_per_token"] == expected_output
assert info["cache_read_input_token_cost"] == expected_cache
assert info["max_input_tokens"] == 1_000_000
assert info.get("supports_vision", False) is expected_vision
# --- provider-prefixed names ---
for key, expected_input, expected_output, expected_cache in [
("deepseek/deepseek-v4-flash", 3e-07, 1.2e-06, 6e-09),
("deepseek/deepseek-v4-pro", 1.32e-06, 3.96e-06, 4.4e-08),
for key, expected_input, expected_output, expected_cache, expected_vision in [
("deepseek/deepseek-v4-flash", 3e-07, 1.2e-06, 6e-09, True),
("deepseek/deepseek-v4-pro", 1.32e-06, 3.96e-06, 4.4e-08, False),
]:
info = model_cost.get(key)
assert info is not None, f"{key} missing from backup JSON"
@ -4157,6 +4174,21 @@ def test_deepseek_v4_models_in_backup_cost_map():
assert info["input_cost_per_token"] == expected_input
assert info["output_cost_per_token"] == expected_output
assert info["cache_read_input_token_cost"] == expected_cache
assert info.get("supports_vision", False) is expected_vision
def test_deprecation_dates_for_retired_xai_and_groq_models():
import json
from pathlib import Path
json_path = Path(__file__).parents[2] / "model_prices_and_context_window.json"
with open(json_path) as f:
model_cost = json.load(f)
assert model_cost["xai/grok-imagine-image-quality"]["deprecation_date"] == "2026-11-02"
assert model_cost["xai/grok-imagine-image-quality-latest"]["deprecation_date"] == "2026-11-02"
assert model_cost["xai/grok-imagine-image-quality-20260403"]["deprecation_date"] == "2026-11-02"
assert model_cost["groq/gemma-7b-it"]["deprecation_date"] == "2024-12-18"
@pytest.mark.usefixtures("local_model_cost_map")