scriptoriumadmin/litellm
1
0
Fork 0
mirror of https://github.com/BerriAI/litellm.git synced 2026-09-14 23:21:35 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 102

Merge c59b321621 into 0c98afa780

Browse source
This commit is contained in:
curiousbox 2026-09-12 14:54:23 -04:00 • committed by GitHub
commit 84f318bb91
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
12 changed files with 591 additions and 0 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
litellm/constants.py
View file

@ -895,6 +895,7 @@ openai_compatible_endpoints: Final[list] = [
"https://api.cognition.ai/v1",
"https://api.scx.ai/v1",
"https://gigachat.devices.sberbank.ru/api/v1",
"https://api.tium.ai/v1",
]
@ -966,6 +967,7 @@ openai_compatible_providers: Final[list] = [
"meta", # Meta Model API (Muse Spark) - JSON-configured provider
"cognition",
"scx-ai",
"tium",
]
openai_text_completion_compatible_providers: Final[list] = [ # providers that support `/v1/completions`
"together_ai",

11
litellm/llms/openai_like/providers.json
View file

@ -200,5 +200,16 @@
"temperature_max": 1.99
},
"supported_endpoints": ["/v1/chat/completions"]
},
"tium": {
"base_url": "https://api.tium.ai/v1",
"api_key_env": "TIUM_API_KEY",
"api_base_env": "TIUM_API_BASE",
"param_mappings": {
"max_completion_tokens": "max_tokens"
},
"supported_endpoints": [
"/v1/chat/completions"
]
}
}

85
litellm/model_prices_and_context_window_backup.json
View file

@ -43130,6 +43130,91 @@
"output_cost_per_token": 2.8e-05,
"source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models"
},
"tium/glm-5.3-flash": {
"cache_read_input_token_cost": 3.03e-08,
"input_cost_per_token": 1.52e-07,
"litellm_provider": "tium",
"max_input_tokens": 128000,
"max_output_tokens": 32768,
"max_tokens": 32768,
"mode": "chat",
"output_cost_per_token": 5.05e-07,
"source": "https://tium.ai/pricing",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true
},
"tium/deepseek-v4-flash": {
"cache_read_input_token_cost": 2.83e-08,
"input_cost_per_token": 8.89e-07,
"litellm_provider": "tium",
"max_input_tokens": 128000,
"max_output_tokens": 32768,
"max_tokens": 32768,
"mode": "chat",
"output_cost_per_token": 2.668e-06,
"source": "https://tium.ai/pricing",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": false,
"supports_tool_choice": true,
"supports_vision": false
},
"tium/deepseek-v4-pro": {
"cache_read_input_token_cost": 8.89e-08,
"input_cost_per_token": 2.668e-06,
"litellm_provider": "tium",
"max_input_tokens": 128000,
"max_output_tokens": 32768,
"max_tokens": 32768,
"mode": "chat",
"output_cost_per_token": 8.005e-06,
"source": "https://tium.ai/pricing",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": false,
"supports_tool_choice": true,
"supports_vision": false
},
"tium/glm-5.3": {
"cache_read_input_token_cost": 5.26e-07,
"input_cost_per_token": 2.83e-06,
"litellm_provider": "tium",
"max_input_tokens": 128000,
"max_output_tokens": 32768,
"max_tokens": 32768,
"mode": "chat",
"output_cost_per_token": 8.895e-06,
"source": "https://tium.ai/pricing",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": false
},
"tium/kimi-k3": {
"cache_read_input_token_cost": 6.06e-07,
"input_cost_per_token": 6.065e-06,
"litellm_provider": "tium",
"max_input_tokens": 128000,
"max_output_tokens": 32768,
"max_tokens": 32768,
"mode": "chat",
"output_cost_per_token": 3.0323e-05,
"source": "https://tium.ai/pricing",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true
},
"together-ai-21.1b-41b": {
"input_cost_per_token": 8e-07,
"litellm_provider": "together_ai",

17
litellm/provider_endpoints_support_backup.json
View file

@ -2152,6 +2152,23 @@
"interactions": true
}
},
"tium": {
"display_name": "Tium (`tium`)",
"url": "https://docs.litellm.ai/docs/providers/tium",
"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,
"a2a": false
}
},
"together_ai": {
"display_name": "Together AI (`together_ai`)",
"url": "https://docs.litellm.ai/docs/providers/togetherai",

28
litellm/proxy/public_endpoints/provider_create_fields.json
View file

@ -2953,6 +2953,34 @@
],
"default_model_placeholder": "gpt-3.5-turbo"
},
{
"provider": "TIUM",
"provider_display_name": "Tium",
"litellm_provider": "tium",
"credential_fields": [
{
"key": "api_base",
"label": "API Base",
"placeholder": "https://api.tium.ai/v1",
"tooltip": null,
"required": false,
"field_type": "text",
"options": null,
"default_value": null
},
{
"key": "api_key",
"label": "API Key",
"placeholder": null,
"tooltip": null,
"required": true,
"field_type": "password",
"options": null,
"default_value": null
}
],
"default_model_placeholder": "tium/glm-5.3"
},
{
"provider": "TogetherAI",
"provider_display_name": "TogetherAI",

1
litellm/types/utils.py
View file

@ -4012,6 +4012,7 @@ class LlmProviders(str, Enum):
PINSTRIPES = "pinstripes"
COGNITION = "cognition"
SCX_AI = "scx-ai"
TIUM = "tium"
DARKBLOOM = "darkbloom"
META = "meta"
LITELLM_AGENT = "litellm_agent"

85
model_prices_and_context_window.json
View file

@ -43130,6 +43130,91 @@
"output_cost_per_token": 2.8e-05,
"source": "https://cloud.google.com/vertex-ai/generative-ai/docs/learn/models#foundation_models"
},
"tium/glm-5.3-flash": {
"cache_read_input_token_cost": 3.03e-08,
"input_cost_per_token": 1.52e-07,
"litellm_provider": "tium",
"max_input_tokens": 128000,
"max_output_tokens": 32768,
"max_tokens": 32768,
"mode": "chat",
"output_cost_per_token": 5.05e-07,
"source": "https://tium.ai/pricing",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true
},
"tium/deepseek-v4-flash": {
"cache_read_input_token_cost": 2.83e-08,
"input_cost_per_token": 8.89e-07,
"litellm_provider": "tium",
"max_input_tokens": 128000,
"max_output_tokens": 32768,
"max_tokens": 32768,
"mode": "chat",
"output_cost_per_token": 2.668e-06,
"source": "https://tium.ai/pricing",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": false,
"supports_tool_choice": true,
"supports_vision": false
},
"tium/deepseek-v4-pro": {
"cache_read_input_token_cost": 8.89e-08,
"input_cost_per_token": 2.668e-06,
"litellm_provider": "tium",
"max_input_tokens": 128000,
"max_output_tokens": 32768,
"max_tokens": 32768,
"mode": "chat",
"output_cost_per_token": 8.005e-06,
"source": "https://tium.ai/pricing",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": false,
"supports_tool_choice": true,
"supports_vision": false
},
"tium/glm-5.3": {
"cache_read_input_token_cost": 5.26e-07,
"input_cost_per_token": 2.83e-06,
"litellm_provider": "tium",
"max_input_tokens": 128000,
"max_output_tokens": 32768,
"max_tokens": 32768,
"mode": "chat",
"output_cost_per_token": 8.895e-06,
"source": "https://tium.ai/pricing",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": false
},
"tium/kimi-k3": {
"cache_read_input_token_cost": 6.06e-07,
"input_cost_per_token": 6.065e-06,
"litellm_provider": "tium",
"max_input_tokens": 128000,
"max_output_tokens": 32768,
"max_tokens": 32768,
"mode": "chat",
"output_cost_per_token": 3.0323e-05,
"source": "https://tium.ai/pricing",
"supports_function_calling": true,
"supports_prompt_caching": true,
"supports_reasoning": true,
"supports_response_schema": true,
"supports_tool_choice": true,
"supports_vision": true
},
"together-ai-21.1b-41b": {
"input_cost_per_token": 8e-07,
"litellm_provider": "together_ai",

17
provider_endpoints_support.json
View file

@ -2438,6 +2438,23 @@
"interactions": true
}
},
"tium": {
"display_name": "Tium (`tium`)",
"url": "https://docs.litellm.ai/docs/providers/tium",
"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,
"a2a": false
}
},
"together_ai": {
"display_name": "Together AI (`together_ai`)",
"url": "https://docs.litellm.ai/docs/providers/togetherai",

321
tests/test_litellm/llms/openai_like/test_tium_provider.py Normal file
View file

@ -0,0 +1,321 @@
"""
Tests for Tium provider configuration and integration.
"""
import json
import httpx
import pytest
import respx
import litellm
from litellm import completion
TIUM_URL = "https://api.tium.ai/v1/chat/completions"
CHAT_PAYLOAD = {
"id": "chatcmpl-1",
"object": "chat.completion",
"created": 1,
"model": "glm-5.3",
"choices": [
{
"index": 0,
"message": {"role": "assistant", "content": "ok"},
"finish_reason": "stop",
}
],
"usage": {"prompt_tokens": 5, "completion_tokens": 1, "total_tokens": 6},
}
TOOL_PAYLOAD = {
"id": "chatcmpl-2",
"object": "chat.completion",
"created": 1,
"model": "glm-5.3",
"choices": [
{
"index": 0,
"message": {
"role": "assistant",
"content": None,
"tool_calls": [
{
"id": "call_1",
"type": "function",
"function": {
"name": "get_weather",
"arguments": '{"city": "Berlin"}',
},
}
],
},
"finish_reason": "tool_calls",
}
],
"usage": {"prompt_tokens": 5, "completion_tokens": 1, "total_tokens": 6},
}
WEATHER_TOOL = {
"type": "function",
"function": {
"name": "get_weather",
"description": "Get the current weather in a location",
"parameters": {
"type": "object",
"properties": {"city": {"type": "string"}},
"required": ["city"],
},
},
}
class TestTiumProviderConfig:
def test_tium_in_provider_list(self):
from litellm import LlmProviders
assert hasattr(LlmProviders, "TIUM")
assert LlmProviders.TIUM.value == "tium"
assert "tium" in litellm.provider_list
def test_tium_json_config_exists(self):
from litellm.llms.openai_like.json_loader import JSONProviderRegistry
assert JSONProviderRegistry.exists("tium")
tium = JSONProviderRegistry.get("tium")
assert tium is not None
assert tium.base_url == "https://api.tium.ai/v1"
assert tium.api_key_env == "TIUM_API_KEY"
assert tium.api_base_env == "TIUM_API_BASE"
assert tium.param_mappings.get("max_completion_tokens") == "max_tokens"
def test_tium_in_openai_compatible_providers(self):
from litellm.constants import openai_compatible_providers
assert "tium" in openai_compatible_providers
def test_tium_provider_resolution(self):
from litellm.litellm_core_utils.get_llm_provider_logic import get_llm_provider
model, provider, api_key, api_base = get_llm_provider(
model="tium/glm-5.3",
custom_llm_provider=None,
api_base=None,
api_key=None,
)
assert model == "glm-5.3"
assert provider == "tium"
assert api_base == "https://api.tium.ai/v1"
def test_tium_api_base_override(self):
from litellm.litellm_core_utils.get_llm_provider_logic import get_llm_provider
model, provider, api_key, api_base = get_llm_provider(
model="tium/glm-5.3",
custom_llm_provider=None,
api_base="https://custom.example.com/v1",
api_key="sk-test",
)
assert provider == "tium"
assert api_base == "https://custom.example.com/v1"
assert api_key == "sk-test"
def test_tium_url_autodetection(self):
from litellm.litellm_core_utils.get_llm_provider_logic import get_llm_provider
model, provider, api_key, api_base = get_llm_provider(
model="glm-5.3",
custom_llm_provider=None,
api_base="https://api.tium.ai/v1",
api_key=None,
)
assert provider == "tium"
assert api_base == "https://api.tium.ai/v1"
def test_tium_router_config(self):
from litellm import Router
router = Router(
model_list=[
{
"model_name": "tium-chat",
"litellm_params": {
"model": "tium/glm-5.3",
"api_key": "test-key",
},
}
]
)
assert len(router.model_list) == 1
assert router.model_list[0]["model_name"] == "tium-chat"
@pytest.mark.usefixtures("local_model_cost_map")
class TestTiumCompletion:
@respx.mock
def test_tium_completion_hits_tium_url_with_bearer_auth(self):
route = respx.post(TIUM_URL).mock(
return_value=httpx.Response(200, json=CHAT_PAYLOAD)
)
response = completion(
model="tium/glm-5.3",
messages=[{"role": "user", "content": "Hello"}],
api_key="sk-tium-test",
)
assert route.called
request = route.calls.last.request
assert str(request.url) == TIUM_URL
assert request.headers["authorization"] == "Bearer sk-tium-test"
body = json.loads(request.content)
assert body["model"] == "glm-5.3"
assert body["messages"] == [{"role": "user", "content": "Hello"}]
assert response.choices[0].message.content == "ok"
@respx.mock
def test_tium_completion_honours_api_base_override(self):
route = respx.post("https://custom.example.com/v1/chat/completions").mock(
return_value=httpx.Response(200, json=CHAT_PAYLOAD)
)
completion(
model="tium/glm-5.3",
messages=[{"role": "user", "content": "Hello"}],
api_key="sk-tium-test",
api_base="https://custom.example.com/v1",
)
assert route.called
@respx.mock
def test_tium_completion_sends_tools_and_returns_tool_calls(self):
route = respx.post(TIUM_URL).mock(
return_value=httpx.Response(200, json=TOOL_PAYLOAD)
)
response = completion(
model="tium/glm-5.3",
messages=[{"role": "user", "content": "Weather in Berlin?"}],
tools=[WEATHER_TOOL],
tool_choice="auto",
api_key="sk-tium-test",
)
assert route.called
body = json.loads(route.calls.last.request.content)
assert body["tools"] == [WEATHER_TOOL]
assert body["tool_choice"] == "auto"
tool_calls = response.choices[0].message.tool_calls
assert tool_calls is not None
assert tool_calls[0].function.name == "get_weather"
@respx.mock
def test_tium_completion_maps_max_completion_tokens_on_the_wire(self):
route = respx.post(TIUM_URL).mock(
return_value=httpx.Response(200, json=CHAT_PAYLOAD)
)
completion(
model="tium/glm-5.3",
messages=[{"role": "user", "content": "Hello"}],
max_completion_tokens=256,
api_key="sk-tium-test",
)
assert route.called
body = json.loads(route.calls.last.request.content)
assert body["max_tokens"] == 256
assert "max_completion_tokens" not in body
class TestTiumModelMetadata:
TIUM_MODELS = (
"tium/glm-5.3-flash",
"tium/deepseek-v4-flash",
"tium/deepseek-v4-pro",
"tium/glm-5.3",
"tium/kimi-k3",
)
VISION_MODELS = ("tium/glm-5.3-flash", "tium/kimi-k3")
NO_RESPONSE_SCHEMA = ("tium/deepseek-v4-flash", "tium/deepseek-v4-pro")
@staticmethod
def _load(path_parts):
from pathlib import Path
json_path = Path(__file__).parents[4].joinpath(*path_parts)
with open(json_path) as f:
return json.load(f)
def test_tium_models_registered_with_correct_metadata(self):
model_cost = self._load(("model_prices_and_context_window.json",))
for model in self.TIUM_MODELS:
info = model_cost.get(model)
assert info is not None, f"{model} missing from model_prices_and_context_window.json"
assert info["litellm_provider"] == "tium"
assert info["mode"] == "chat"
assert info["input_cost_per_token"] > 0
assert info["output_cost_per_token"] > 0
assert info["supports_function_calling"] is True
assert info["supports_tool_choice"] is True
assert info["supports_reasoning"] is True
assert info["supports_response_schema"] is (model not in self.NO_RESPONSE_SCHEMA)
assert info.get("supports_vision", False) is (model in self.VISION_MODELS)
assert info["supports_prompt_caching"] is True
assert 0 < info["cache_read_input_token_cost"] < info["input_cost_per_token"]
assert info["max_input_tokens"] == 128000
assert info["max_output_tokens"] == 32768
assert info["max_tokens"] == info["max_output_tokens"]
def test_tium_models_synced_to_backup(self):
model_cost = self._load(("model_prices_and_context_window.json",))
backup = self._load(("litellm", "model_prices_and_context_window_backup.json"))
for model in self.TIUM_MODELS:
assert model in backup, f"{model} missing from backup json"
assert backup[model] == model_cost[model], f"{model} differs between root and backup json"
def test_tium_supported_endpoints_matrix(self):
from pathlib import Path
import litellm as _litellm
backup_path = (
Path(_litellm.__file__).parent / "provider_endpoints_support_backup.json"
)
matrix = json.loads(backup_path.read_text())
assert "tium" in matrix["providers"]
endpoints = matrix["providers"]["tium"]["endpoints"]
assert endpoints["chat_completions"] is True
assert endpoints["responses"] is False
assert endpoints["embeddings"] is False
def test_tium_listed_in_add_model_form(self):
entries = self._load(
("litellm", "proxy", "public_endpoints", "provider_create_fields.json")
)
tium = [e for e in entries if e["litellm_provider"] == "tium"]
assert len(tium) == 1
entry = tium[0]
assert entry["provider"] == "TIUM"
assert entry["provider_display_name"] == "Tium"
assert entry["default_model_placeholder"].startswith("tium/")
fields = {f["key"]: f for f in entry["credential_fields"]}
assert fields["api_key"]["required"] is True
assert fields["api_key"]["field_type"] == "password"
assert fields["api_base"]["required"] is False

4
ui/litellm-dashboard/public/assets/logos/tium.svg Normal file
View file

@ -0,0 +1,4 @@
<svg xmlns="http://www.w3.org/2000/svg" width="32" height="32" viewBox="0 0 32 32">
<rect width="32" height="32" fill="#08090A"/>
<rect x="10" y="6" width="12" height="20" rx="1" fill="#FF9E2C"/>
</svg>
Side by side

After

Width:  |  Height:  |  Size: 207 B

15
ui/litellm-dashboard/src/components/provider_info_helpers.test.tsx
View file

@ -73,6 +73,17 @@ describe("provider_info_helpers", () => {
expect(fromEnumKey.logo).toBe(providerLogoMap[Providers.SCX_AI]);
});
it("should map tium slug and TIUM enum key to the Tium display name and logo", () => {
const fromSlug = getProviderLogoAndName("tium");
expect(fromSlug.displayName).toBe(Providers.TIUM);
expect(fromSlug.logo).toBe(providerLogoMap[Providers.TIUM]);
expect(fromSlug.logo).toBeTruthy();
const fromEnumKey = getProviderLogoAndName("TIUM");
expect(fromEnumKey.displayName).toBe(Providers.TIUM);
expect(fromEnumKey.logo).toBe(providerLogoMap[Providers.TIUM]);
});
it("should map bedrock_mantle slug to Bedrock Mantle display name and logo", () => {
const result = getProviderLogoAndName("bedrock_mantle");
expect(result.displayName).toBe(Providers.BedrockMantle);
@ -205,6 +216,10 @@ describe("provider_info_helpers", () => {
expect(getPlaceholder(Providers.SCX_AI)).toBe("scx-ai/GLM-5.2");
});
it("should return a tium model placeholder for TIUM provider", () => {
expect(getPlaceholder(Providers.TIUM)).toBe("tium/glm-5.3");
});
it("should return claude-3-opus placeholder for Anthropic provider", () => {
expect(getPlaceholder(Providers.Anthropic)).toBe("claude-3-opus");
});

5
ui/litellm-dashboard/src/components/provider_info_helpers.tsx
View file

@ -54,6 +54,7 @@ import sapLogo from "../../public/assets/logos/sap.png";
import scxAiLogo from "../../public/assets/logos/scx_ai.svg";
import snowflakeLogo from "../../public/assets/logos/snowflake.svg";
import sonioxLogo from "../../public/assets/logos/soniox.svg";
import tiumLogo from "../../public/assets/logos/tium.svg";
import togetheraiLogo from "../../public/assets/logos/togetherai.svg";
import topazLogo from "../../public/assets/logos/topaz.svg";
import v0Logo from "../../public/assets/logos/v0.svg";
@ -163,6 +164,7 @@ export enum Providers {
Snowflake = "Snowflake",
Soniox = "Soniox",
TEXT_COMPLETION_CODESTRAL = "Text-Completion-Codestral",
TIUM = "Tium",
TogetherAI = "TogetherAI",
TOPAZ = "Topaz",
Triton = "Triton",
@ -279,6 +281,7 @@ export const provider_map: Record<string, string> = {
Snowflake: "snowflake",
Soniox: "soniox",
TEXT_COMPLETION_CODESTRAL: "text-completion-codestral",
TIUM: "tium",
TogetherAI: "together_ai",
TOPAZ: "topaz",
Triton: "triton",
@ -376,6 +379,7 @@ export const providerLogoMap: Partial<Record<Providers, string>> = {
[Providers.Snowflake]: snowflakeLogo.src,
[Providers.Soniox]: sonioxLogo.src,
[Providers.TEXT_COMPLETION_CODESTRAL]: mistralLogo.src,
[Providers.TIUM]: tiumLogo.src,
[Providers.TogetherAI]: togetheraiLogo.src,
[Providers.TOPAZ]: topazLogo.src,
[Providers.Triton]: nvidiaTritonLogo.src,
@ -441,6 +445,7 @@ const providerPlaceholderMap: Partial<Record<Providers, string>> = {
[Providers.SageMaker]: "sagemaker/jumpstart-dft-meta-textgeneration-llama-2-7b",
[Providers.SCX_AI]: "scx-ai/GLM-5.2",
[Providers.Snowflake]: "snowflake/mistral-7b",
[Providers.TIUM]: "tium/glm-5.3",
[Providers.Vertex_AI]: "gemini-pro",
[Providers.VolcEngine]: "volcengine/<any-model-on-volcengine>",
[Providers.Voyage]: "voyage/",