This commit is contained in:
Christian Landgren 2026-09-21 04:54:48 -04:00 • committed by GitHub
commit 697bf5c5b6
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
5 changed files with 280 additions and 0 deletions

View file

@ -200,5 +200,10 @@
"temperature_max": 1.99
},
"supported_endpoints": ["/v1/chat/completions"]
},
"berget": {
"base_url": "https://api.berget.ai/v1",
"api_key_env": "BERGET_API_KEY",
"api_base_env": "BERGET_API_BASE"
}
}

View file

@ -75170,5 +75170,74 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_web_search": false
},
"berget/Qwen/Qwen3.8-27B-FP8": {
"litellm_provider": "berget",
"mode": "chat",
"input_cost_per_token": 4.6000000000000004e-07,
"output_cost_per_token": 3.44e-06,
"max_input_tokens": 262144,
"max_output_tokens": 262144,
"max_tokens": 262144,
"supports_function_calling": true,
"supports_vision": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_reasoning": true
},
"berget/google/gemma-4-31B-it": {
"litellm_provider": "berget",
"mode": "chat",
"input_cost_per_token": 2.9e-07,
"output_cost_per_token": 5.699999999999999e-07,
"max_input_tokens": 262144,
"max_output_tokens": 262144,
"max_tokens": 262144,
"supports_function_calling": true,
"supports_vision": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_reasoning": true
},
"berget/mistralai/Mistral-Small-3.2-24B-Instruct-2506": {
"litellm_provider": "berget",
"mode": "chat",
"input_cost_per_token": 3.4000000000000003e-07,
"output_cost_per_token": 3.4000000000000003e-07,
"max_input_tokens": 131072,
"max_output_tokens": 131072,
"max_tokens": 131072,
"supports_function_calling": true,
"supports_vision": false,
"supports_response_schema": true,
"supports_system_messages": true
},
"berget/moonshotai/Kimi-K3": {
"litellm_provider": "berget",
"mode": "chat",
"input_cost_per_token": 3.44e-06,
"output_cost_per_token": 1.7219999999999998e-05,
"max_input_tokens": 327680,
"max_output_tokens": 327680,
"max_tokens": 327680,
"supports_function_calling": true,
"supports_vision": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_reasoning": true
},
"berget/zai-org/GLM-5.3-Flash": {
"litellm_provider": "berget",
"mode": "chat",
"input_cost_per_token": 2.9e-07,
"output_cost_per_token": 5.699999999999999e-07,
"max_input_tokens": 1048576,
"max_output_tokens": 1048576,
"max_tokens": 1048576,
"supports_function_calling": true,
"supports_vision": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_reasoning": true
}
}

View file

@ -75170,5 +75170,74 @@
"supports_tool_choice": true,
"supports_vision": true,
"supports_web_search": false
},
"berget/Qwen/Qwen3.8-27B-FP8": {
"litellm_provider": "berget",
"mode": "chat",
"input_cost_per_token": 4.6000000000000004e-07,
"output_cost_per_token": 3.44e-06,
"max_input_tokens": 262144,
"max_output_tokens": 262144,
"max_tokens": 262144,
"supports_function_calling": true,
"supports_vision": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_reasoning": true
},
"berget/google/gemma-4-31B-it": {
"litellm_provider": "berget",
"mode": "chat",
"input_cost_per_token": 2.9e-07,
"output_cost_per_token": 5.699999999999999e-07,
"max_input_tokens": 262144,
"max_output_tokens": 262144,
"max_tokens": 262144,
"supports_function_calling": true,
"supports_vision": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_reasoning": true
},
"berget/mistralai/Mistral-Small-3.2-24B-Instruct-2506": {
"litellm_provider": "berget",
"mode": "chat",
"input_cost_per_token": 3.4000000000000003e-07,
"output_cost_per_token": 3.4000000000000003e-07,
"max_input_tokens": 131072,
"max_output_tokens": 131072,
"max_tokens": 131072,
"supports_function_calling": true,
"supports_vision": false,
"supports_response_schema": true,
"supports_system_messages": true
},
"berget/moonshotai/Kimi-K3": {
"litellm_provider": "berget",
"mode": "chat",
"input_cost_per_token": 3.44e-06,
"output_cost_per_token": 1.7219999999999998e-05,
"max_input_tokens": 327680,
"max_output_tokens": 327680,
"max_tokens": 327680,
"supports_function_calling": true,
"supports_vision": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_reasoning": true
},
"berget/zai-org/GLM-5.3-Flash": {
"litellm_provider": "berget",
"mode": "chat",
"input_cost_per_token": 2.9e-07,
"output_cost_per_token": 5.699999999999999e-07,
"max_input_tokens": 1048576,
"max_output_tokens": 1048576,
"max_tokens": 1048576,
"supports_function_calling": true,
"supports_vision": true,
"supports_response_schema": true,
"supports_system_messages": true,
"supports_reasoning": true
}
}

View file

@ -2967,6 +2967,35 @@
"a2a": false
}
},
"berget": {
"display_name": "Berget (`berget`)",
"url": "https://www.berget.ai/",
"endpoints": {
"chat_completions": true,
"messages": false,
"responses": false,
"embeddings": false,
"image_generations": false,
"audio_transcriptions": false,
"audio_speech": false,
"moderations": false,
"batches": false,
"rerank": false,
"ocr": false,
"search": false,
"skills": false,
"interactions": false,
"a2a_(Agent Gateway)": false,
"container": false,
"container_file": false,
"compact": false,
"files": false,
"image_edits": false,
"vector_stores_create": false,
"vector_stores_search": false,
"video_generations": false
}
},
"searchapi": {
"display_name": "SearchAPI (`searchapi`)",
"url": "https://docs.litellm.ai/docs/providers/searchapi",

View file

@ -0,0 +1,108 @@
import json
import os
from pathlib import Path
try:
import pytest
except ImportError:
pytest = None
import litellm
from litellm.llms.openai_like.dynamic_config import create_config_class
from litellm.llms.openai_like.json_loader import JSONProviderRegistry
REPO_ROOT = Path(__file__).resolve().parents[4]
BERGET_MODELS = [
"Qwen/Qwen3.8-27B-FP8",
"google/gemma-4-31B-it",
"mistralai/Mistral-Small-3.2-24B-Instruct-2506",
"moonshotai/Kimi-K3",
"zai-org/GLM-5.3-Flash",
]
class TestBergetProvider:
def test_provider_registered(self):
assert JSONProviderRegistry.exists("berget")
def test_provider_config(self):
provider = JSONProviderRegistry.get("berget")
assert provider.base_url == "https://api.berget.ai/v1"
assert provider.api_key_env == "BERGET_API_KEY"
assert provider.api_base_env == "BERGET_API_BASE"
def test_dynamic_config_resolves_api_base(self):
provider = JSONProviderRegistry.get("berget")
config = create_config_class(provider)()
api_base, _ = config._get_openai_compatible_provider_info(None, None)
assert api_base == "https://api.berget.ai/v1"
def test_dynamic_config_custom_overrides(self):
provider = JSONProviderRegistry.get("berget")
config = create_config_class(provider)()
api_base, api_key = config._get_openai_compatible_provider_info(
"https://api-stage.berget.ai/v1", "test-key"
)
assert api_base == "https://api-stage.berget.ai/v1"
assert api_key == "test-key"
def test_provider_inference_resolves_prefix(self):
from litellm.litellm_core_utils.get_llm_provider_logic import get_llm_provider
model, provider, _, api_base = get_llm_provider("berget/zai-org/GLM-5.3-Flash")
assert provider == "berget"
assert model == "zai-org/GLM-5.3-Flash"
assert api_base == "https://api.berget.ai/v1"
def test_no_param_mappings_required_for_berget(self):
provider = JSONProviderRegistry.get("berget")
assert not getattr(provider, "param_mappings", {})
def test_provider_support_documented(self):
support = json.loads(
(REPO_ROOT / "provider_endpoints_support.json").read_text()
)
assert support["providers"]["berget"]["endpoints"]["chat_completions"] is True
assert support["providers"]["berget"]["endpoints"]["responses"] is False
@pytest.mark.skipif(pytest is None, reason="pytest not installed")
@pytest.mark.parametrize("model_id", BERGET_MODELS)
def test_pricing_entries_exist(self, model_id):
entries = json.loads(
(REPO_ROOT / "model_prices_and_context_window.json").read_text()
)
key = f"berget/{model_id}"
assert key in entries, key
entry = entries[key]
assert entry["litellm_provider"] == "berget"
assert entry["mode"] == "chat"
assert entry["input_cost_per_token"] > 0
assert entry["output_cost_per_token"] > 0
assert entry["max_input_tokens"] > 0
@pytest.mark.skipif(pytest is None, reason="pytest not installed")
def test_runtime_backup_matches_root_cost_map(self):
root = json.loads(
(REPO_ROOT / "model_prices_and_context_window.json").read_text()
)
backup = json.loads(
(REPO_ROOT / "litellm" / "model_prices_and_context_window_backup.json").read_text()
)
assert {k: v for k, v in root.items() if k.startswith("berget")} == {
k: v for k, v in backup.items() if k.startswith("berget")
}
@pytest.mark.skipif(pytest is None, reason="pytest not installed")
def test_litellm_packaged_backup_used_by_runtime_env(self):
packaged = Path(litellm.__file__).parent
if (packaged / "model_prices_and_context_window_backup.json").is_file():
backup = json.loads(
(packaged / "model_prices_and_context_window_backup.json").read_text()
)
root = json.loads(
(REPO_ROOT / "model_prices_and_context_window.json").read_text()
)
assert {k: v for k, v in root.items() if k.startswith("berget")} == {
k: v for k, v in backup.items() if k.startswith("berget")
}