From 64be726beecbaea1370e042476b6a506dbf72249 Mon Sep 17 00:00:00 2001 From: AJ-ing Date: Thu, 1 Oct 2026 10:05:44 +1000 Subject: [PATCH] feat(openinfer): add OpenInfer as OpenAI-compatible provider Register OpenInfer in the JSON provider registry with catalog models, proxy create fields, and unit coverage so requests route via openinfer/ Co-authored-by: Cursor --- litellm/constants.py | 2 + litellm/llms/openai_like/providers.json | 11 ++ ...odel_prices_and_context_window_backup.json | 44 +++++ .../provider_endpoints_support_backup.json | 17 ++ .../provider_create_fields.json | 28 +++ litellm/types/utils.py | 1 + model_prices_and_context_window.json | 44 +++++ provider_endpoints_support.json | 17 ++ .../openai_like/test_openinfer_provider.py | 173 ++++++++++++++++++ 9 files changed, 337 insertions(+) create mode 100644 tests/unit/llms/openai_like/test_openinfer_provider.py diff --git a/litellm/constants.py b/litellm/constants.py index 9af40744896..43c7a012a78 100644 --- a/litellm/constants.py +++ b/litellm/constants.py @@ -964,6 +964,7 @@ openai_compatible_endpoints: Final[list] = [ "https://api.scx.ai/v1", "https://api.prisminference.com/v1", "https://gigachat.devices.sberbank.ru/api/v1", + "https://api.openinfer.ai/v1", ] @@ -1032,6 +1033,7 @@ openai_compatible_providers: Final[list] = [ "docker_model_runner", "ragflow", "pinstripes", # Pinstripes - JSON-configured provider + "openinfer", "darkbloom", "meta", # Meta Model API (Muse Spark) - JSON-configured provider "cognition", diff --git a/litellm/llms/openai_like/providers.json b/litellm/llms/openai_like/providers.json index 440e5490d71..18c961ecab5 100644 --- a/litellm/llms/openai_like/providers.json +++ b/litellm/llms/openai_like/providers.json @@ -212,5 +212,16 @@ "api_key_env": "SAIL_API_KEY", "api_base_env": "SAIL_API_BASE", "supported_endpoints": ["/v1/chat/completions", "/v1/responses", "/v1/messages"] + }, + "openinfer": { + "base_url": "https://api.openinfer.ai/v1", + "api_key_env": "OPENINFER_API_KEY", + "api_base_env": "OPENINFER_API_BASE", + "base_class": "openai_gpt", + "param_mappings": { + "max_completion_tokens": "max_tokens" + }, + "supported_endpoints": ["/v1/chat/completions"] } } + diff --git a/litellm/model_prices_and_context_window_backup.json b/litellm/model_prices_and_context_window_backup.json index 136399b557e..2bb21b5eea8 100644 --- a/litellm/model_prices_and_context_window_backup.json +++ b/litellm/model_prices_and_context_window_backup.json @@ -79420,5 +79420,49 @@ "supported_endpoints": [ "/v1/audio/speech" ] + }, + "openinfer/@oi/Llama-3.2-1B-Instruct": { + "max_tokens": 8192, + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-08, + "output_cost_per_token": 2e-08, + "litellm_provider": "openinfer", + "mode": "chat", + "supports_function_calling": true, + "source": "https://api.openinfer.ai/v1/models" + }, + "openinfer/@oi/Qwen3.5-9B": { + "max_tokens": 8192, + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 1.8e-07, + "litellm_provider": "openinfer", + "mode": "chat", + "supports_function_calling": true, + "source": "https://api.openinfer.ai/v1/models" + }, + "openinfer/@oi/Qwen3.5-27B": { + "max_tokens": 8192, + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "input_cost_per_token": 7.2e-07, + "output_cost_per_token": 7.2e-07, + "litellm_provider": "openinfer", + "mode": "chat", + "supports_function_calling": true, + "source": "https://api.openinfer.ai/v1/models" + }, + "openinfer/@oi/Gemma4-31B-It": { + "max_tokens": 8192, + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "input_cost_per_token": 5.2e-07, + "output_cost_per_token": 7.5e-07, + "litellm_provider": "openinfer", + "mode": "chat", + "supports_function_calling": true, + "source": "https://api.openinfer.ai/v1/models" } } diff --git a/litellm/provider_endpoints_support_backup.json b/litellm/provider_endpoints_support_backup.json index ad6e5857218..aacdc11a9c2 100644 --- a/litellm/provider_endpoints_support_backup.json +++ b/litellm/provider_endpoints_support_backup.json @@ -2641,6 +2641,23 @@ "messages": true, "responses": true } + }, + "openinfer": { + "display_name": "OpenInfer (`openinfer`)", + "url": "https://openinfer.io", + "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 + } } }, "endpoints": { diff --git a/litellm/proxy/public_endpoints/provider_create_fields.json b/litellm/proxy/public_endpoints/provider_create_fields.json index 67a8c356a4a..a3f62f2c9c2 100644 --- a/litellm/proxy/public_endpoints/provider_create_fields.json +++ b/litellm/proxy/public_endpoints/provider_create_fields.json @@ -2499,6 +2499,34 @@ ], "default_model_placeholder": "gpt-3.5-turbo" }, + { + "provider": "OPENINFER", + "provider_display_name": "OpenInfer", + "litellm_provider": "openinfer", + "credential_fields": [ + { + "key": "api_base", + "label": "API Base", + "placeholder": "https://api.openinfer.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": "openinfer/@oi/Llama-3.2-1B-Instruct" + }, { "provider": "OPENAI_LIKE", "provider_display_name": "Openai Like", diff --git a/litellm/types/utils.py b/litellm/types/utils.py index c12a4def69a..63c587537cf 100644 --- a/litellm/types/utils.py +++ b/litellm/types/utils.py @@ -4155,6 +4155,7 @@ class LlmProviders(str, Enum): COGNITION = "cognition" SCX_AI = "scx-ai" PRISM = "prism" + OPENINFER = "openinfer" DARKBLOOM = "darkbloom" META = "meta" SAIL = "sail" diff --git a/model_prices_and_context_window.json b/model_prices_and_context_window.json index 136399b557e..2bb21b5eea8 100644 --- a/model_prices_and_context_window.json +++ b/model_prices_and_context_window.json @@ -79420,5 +79420,49 @@ "supported_endpoints": [ "/v1/audio/speech" ] + }, + "openinfer/@oi/Llama-3.2-1B-Instruct": { + "max_tokens": 8192, + "max_input_tokens": 128000, + "max_output_tokens": 8192, + "input_cost_per_token": 2e-08, + "output_cost_per_token": 2e-08, + "litellm_provider": "openinfer", + "mode": "chat", + "supports_function_calling": true, + "source": "https://api.openinfer.ai/v1/models" + }, + "openinfer/@oi/Qwen3.5-9B": { + "max_tokens": 8192, + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "input_cost_per_token": 1.5e-07, + "output_cost_per_token": 1.8e-07, + "litellm_provider": "openinfer", + "mode": "chat", + "supports_function_calling": true, + "source": "https://api.openinfer.ai/v1/models" + }, + "openinfer/@oi/Qwen3.5-27B": { + "max_tokens": 8192, + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "input_cost_per_token": 7.2e-07, + "output_cost_per_token": 7.2e-07, + "litellm_provider": "openinfer", + "mode": "chat", + "supports_function_calling": true, + "source": "https://api.openinfer.ai/v1/models" + }, + "openinfer/@oi/Gemma4-31B-It": { + "max_tokens": 8192, + "max_input_tokens": 262144, + "max_output_tokens": 8192, + "input_cost_per_token": 5.2e-07, + "output_cost_per_token": 7.5e-07, + "litellm_provider": "openinfer", + "mode": "chat", + "supports_function_calling": true, + "source": "https://api.openinfer.ai/v1/models" } } diff --git a/provider_endpoints_support.json b/provider_endpoints_support.json index 44ef9363b64..dd25034facc 100644 --- a/provider_endpoints_support.json +++ b/provider_endpoints_support.json @@ -3104,6 +3104,23 @@ "rerank": false, "a2a": false } + }, + "openinfer": { + "display_name": "OpenInfer (`openinfer`)", + "url": "https://openinfer.io", + "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 + } } }, "endpoints": { diff --git a/tests/unit/llms/openai_like/test_openinfer_provider.py b/tests/unit/llms/openai_like/test_openinfer_provider.py new file mode 100644 index 00000000000..0b6cc359113 --- /dev/null +++ b/tests/unit/llms/openai_like/test_openinfer_provider.py @@ -0,0 +1,173 @@ +""" +Tests for the OpenInfer LLM provider configuration and integration. +""" + +import json +from pathlib import Path + +import pytest + +import litellm + +_REPO_ROOT = Path(__file__).resolve().parents[4] +_PRICE_FILES = ( + _REPO_ROOT / "model_prices_and_context_window.json", + _REPO_ROOT / "litellm" / "model_prices_and_context_window_backup.json", +) + + +class TestOpenInferProviderConfig: + def test_openinfer_in_provider_list(self): + from litellm import LlmProviders + + assert LlmProviders.OPENINFER.value == "openinfer" + assert "openinfer" in litellm.provider_list + + def test_openinfer_json_config(self): + from litellm.llms.openai_like.json_loader import JSONProviderRegistry + + provider = JSONProviderRegistry.get("openinfer") + assert provider is not None + assert provider.base_url == "https://api.openinfer.ai/v1" + assert provider.api_key_env == "OPENINFER_API_KEY" + assert provider.api_base_env == "OPENINFER_API_BASE" + assert not JSONProviderRegistry.supports_responses_api("openinfer") + + def test_openinfer_in_openai_compatible_providers(self): + from litellm.constants import openai_compatible_providers + + assert "openinfer" in openai_compatible_providers + + def test_provider_prefixed_model_routes_to_openinfer(self): + from litellm.litellm_core_utils.get_llm_provider_logic import get_llm_provider + + model, provider, api_key, api_base = get_llm_provider( + model="openinfer/@oi/Llama-3.2-1B-Instruct", + custom_llm_provider=None, + api_base=None, + api_key="sk-test", + ) + + assert model == "@oi/Llama-3.2-1B-Instruct" + assert provider == "openinfer" + assert api_key == "sk-test" + assert api_base == "https://api.openinfer.ai/v1" + + def test_api_key_and_base_resolved_from_env(self, monkeypatch): + from litellm.litellm_core_utils.get_llm_provider_logic import get_llm_provider + + monkeypatch.setenv("OPENINFER_API_KEY", "sk-env-key") + monkeypatch.setenv("OPENINFER_API_BASE", "https://proxy.internal/v1") + + _, provider, api_key, api_base = get_llm_provider( + model="openinfer/@oi/Qwen3.5-9B", + custom_llm_provider=None, + api_base=None, + api_key=None, + ) + + assert provider == "openinfer" + assert api_key == "sk-env-key" + assert api_base == "https://proxy.internal/v1" + + def test_url_autodetection_from_api_base(self, monkeypatch): + from litellm.litellm_core_utils.get_llm_provider_logic import get_llm_provider + + monkeypatch.setenv("OPENINFER_API_KEY", "sk-env-key") + + _, provider, api_key, api_base = get_llm_provider( + model="@oi/Llama-3.2-1B-Instruct", + custom_llm_provider=None, + api_base="https://api.openinfer.ai/v1", + api_key=None, + ) + + assert provider == "openinfer" + assert api_key == "sk-env-key" + + def test_chat_completions_url(self): + config = litellm.ProviderConfigManager.get_provider_chat_config( + model="@oi/Llama-3.2-1B-Instruct", provider=litellm.LlmProviders.OPENINFER + ) + assert config is not None + assert ( + config.get_complete_url( + api_base=None, + api_key="sk-test", + model="@oi/Llama-3.2-1B-Instruct", + optional_params={}, + litellm_params={}, + ) + == "https://api.openinfer.ai/v1/chat/completions" + ) + + def test_outgoing_chat_request_contract(self): + from litellm.litellm_core_utils.get_llm_provider_logic import get_llm_provider + from litellm.llms.openai_like.dynamic_config import create_config_class + from litellm.llms.openai_like.json_loader import JSONProviderRegistry + + model, provider, api_key, api_base = get_llm_provider( + model="openinfer/@oi/Llama-3.2-1B-Instruct", + custom_llm_provider=None, + api_base=None, + api_key="sk-test", + ) + assert model == "@oi/Llama-3.2-1B-Instruct" + assert provider == "openinfer" + assert api_key == "sk-test" + assert api_base == "https://api.openinfer.ai/v1" + + provider_cfg = JSONProviderRegistry.get("openinfer") + assert provider_cfg is not None + config = create_config_class(provider_cfg)() + + assert ( + config.get_complete_url( + api_base=None, + api_key=api_key, + model=model, + optional_params={}, + litellm_params={}, + ) + == "https://api.openinfer.ai/v1/chat/completions" + ) + + headers = config.validate_environment( + headers={}, + model=model, + messages=[{"role": "user", "content": "hi"}], + optional_params={}, + litellm_params={}, + api_key=api_key, + api_base=api_base, + ) + assert headers["Authorization"] == "Bearer sk-test" + + optional_params = config.map_openai_params( + non_default_params={"max_completion_tokens": 128}, + optional_params={}, + model=model, + drop_params=False, + ) + assert optional_params["max_tokens"] == 128 + assert "max_completion_tokens" not in optional_params + + @pytest.mark.parametrize( + ("model", "input_cost_per_token", "output_cost_per_token"), + ( + ("openinfer/@oi/Llama-3.2-1B-Instruct", 2e-08, 2e-08), + ("openinfer/@oi/Qwen3.5-9B", 1.5e-07, 1.8e-07), + ("openinfer/@oi/Qwen3.5-27B", 7.2e-07, 7.2e-07), + ("openinfer/@oi/Gemma4-31B-It", 5.2e-07, 7.5e-07), + ), + ) + def test_catalog_token_rates_match_vendor_per_million_prices( + self, model: str, input_cost_per_token: float, output_cost_per_token: float + ): + for path in _PRICE_FILES: + catalog = json.loads(path.read_text()) + row = catalog[model] + assert row["input_cost_per_token"] == input_cost_per_token + assert row["output_cost_per_token"] == output_cost_per_token + assert row["input_cost_per_token"] > 0 + assert row["output_cost_per_token"] > 0