diff --git a/litellm/constants.py b/litellm/constants.py index 2901b7d8948..ffdb661e826 100644 --- a/litellm/constants.py +++ b/litellm/constants.py @@ -1060,6 +1060,7 @@ openai_compatible_endpoints: Final[list[str]] = [ "https://api.scx.ai/v1", "https://api.prisminference.com/v1", "https://api.reka.ai/v1", + "https://api.iteracompute.com/v1", "https://gigachat.devices.sberbank.ru/api/v1", "https://api.openinfer.ai/v1", ] @@ -1141,6 +1142,7 @@ openai_compatible_providers: Final[list[str]] = [ "prism", "sail", "reka", + "iteracompute", ] OPENAI_AUDIO_TRANSCRIPTION_PROVIDERS: Final = frozenset({"openai"} | frozenset(openai_compatible_providers)) diff --git a/litellm/llms/openai_like/providers.json b/litellm/llms/openai_like/providers.json index 115d914d3ac..1427c0b04e8 100644 --- a/litellm/llms/openai_like/providers.json +++ b/litellm/llms/openai_like/providers.json @@ -240,6 +240,11 @@ }, "supported_endpoints": ["/v1/chat/completions", "/v1/responses", "/v1/messages"] }, + "iteracompute": { + "base_url": "https://api.iteracompute.com/v1", + "api_key_env": "ITERACOMPUTE_API_KEY", + "api_base_env": "ITERACOMPUTE_API_BASE" + }, "inferenceapis": { "base_url": "https://api.inferenceapis.com/v1", "api_key_env": "INFERENCEAPIS_API_KEY", diff --git a/litellm/provider_endpoints_support_backup.json b/litellm/provider_endpoints_support_backup.json index 49b667af819..72b8702350e 100644 --- a/litellm/provider_endpoints_support_backup.json +++ b/litellm/provider_endpoints_support_backup.json @@ -2153,6 +2153,24 @@ "interactions": true } }, + "iteracompute": { + "display_name": "IteraCompute (`iteracompute`)", + "url": "https://docs.litellm.ai/docs/providers/iteracompute", + "endpoints": { + "chat_completions": true, + "messages": true, + "responses": true, + "embeddings": false, + "image_generations": false, + "audio_transcriptions": false, + "audio_speech": false, + "moderations": false, + "batches": false, + "rerank": false, + "a2a": false, + "interactions": false + } + }, "reka": { "display_name": "Reka (`reka`)", "url": "https://docs.litellm.ai/docs/providers/reka", diff --git a/litellm/proxy/public_endpoints/provider_create_fields.json b/litellm/proxy/public_endpoints/provider_create_fields.json index d96f2976508..68a586e3dd2 100644 --- a/litellm/proxy/public_endpoints/provider_create_fields.json +++ b/litellm/proxy/public_endpoints/provider_create_fields.json @@ -4144,5 +4144,33 @@ } ], "default_model_placeholder": "cursor/claude-4-sonnet" + }, + { + "provider": "ITERACOMPUTE", + "provider_display_name": "IteraCompute", + "litellm_provider": "iteracompute", + "credential_fields": [ + { + "key": "api_base", + "label": "API Base", + "placeholder": "https://api.iteracompute.com/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": "iteracompute/qwen/qwen3.8-27b" } ] diff --git a/litellm/types/utils.py b/litellm/types/utils.py index 7e8645fd0dd..d25771a462b 100644 --- a/litellm/types/utils.py +++ b/litellm/types/utils.py @@ -4589,6 +4589,7 @@ class LlmProviders(str, Enum): META = "meta" SAIL = "sail" REKA = "reka" + ITERACOMPUTE = "iteracompute" LITELLM_AGENT = "litellm_agent" CURSOR = "cursor" BEDROCK_MANTLE = "bedrock_mantle" diff --git a/provider_endpoints_support.json b/provider_endpoints_support.json index e62b74aa87d..1a44044ae6c 100644 --- a/provider_endpoints_support.json +++ b/provider_endpoints_support.json @@ -3222,6 +3222,24 @@ "rerank": false } }, + "iteracompute": { + "display_name": "IteraCompute (`iteracompute`)", + "url": "https://docs.litellm.ai/docs/providers/iteracompute", + "endpoints": { + "chat_completions": true, + "messages": true, + "responses": true, + "embeddings": false, + "image_generations": false, + "audio_transcriptions": false, + "audio_speech": false, + "moderations": false, + "batches": false, + "rerank": false, + "a2a": false, + "interactions": false + } + }, "reka": { "display_name": "Reka (`reka`)", "url": "https://docs.litellm.ai/docs/providers/reka", diff --git a/tests/unit/llms/openai_like/test_iteracompute_provider.py b/tests/unit/llms/openai_like/test_iteracompute_provider.py new file mode 100644 index 00000000000..039a432d628 --- /dev/null +++ b/tests/unit/llms/openai_like/test_iteracompute_provider.py @@ -0,0 +1,175 @@ +import json +from typing import Final + +import pytest +import respx + +import litellm +from litellm.caching.llm_caching_handler import LLMClientCache +from litellm.litellm_core_utils.get_llm_provider_logic import get_llm_provider + +_COMPLETION: Final = { + "id": "chatcmpl_test", + "object": "chat.completion", + "created": 1_790_000_000, + "model": "qwen/qwen3.8-27b", + "choices": [ + { + "index": 0, + "message": {"role": "assistant", "content": "Hello"}, + "finish_reason": "stop", + } + ], + "usage": {"prompt_tokens": 4, "completion_tokens": 3, "total_tokens": 7}, +} + + +@pytest.mark.parametrize( + "model,custom_provider,api_base", + ( + ("iteracompute/qwen/qwen3.8-27b", None, None), + ("qwen/qwen3.8-27b", "iteracompute", None), + ("qwen/qwen3.8-27b", None, "https://api.iteracompute.com/v1"), + ), +) +def test_provider_resolution_preserves_model_namespace( + monkeypatch: pytest.MonkeyPatch, + model: str, + custom_provider: str | None, + api_base: str | None, +) -> None: + monkeypatch.setenv("ITERACOMPUTE_API_KEY", "test-key") + monkeypatch.delenv("ITERACOMPUTE_API_BASE", raising=False) + + assert get_llm_provider( + model=model, + custom_llm_provider=custom_provider, + api_base=api_base, + api_key=None, + ) == ( + "qwen/qwen3.8-27b", + "iteracompute", + "test-key", + "https://api.iteracompute.com/v1", + ) + + +@pytest.mark.parametrize("explicit", (False, True)) +def test_credentials_respect_explicit_arguments_before_environment( + monkeypatch: pytest.MonkeyPatch, explicit: bool +) -> None: + monkeypatch.setenv("ITERACOMPUTE_API_KEY", "env-key") + monkeypatch.setenv("ITERACOMPUTE_API_BASE", "https://env.example/v1") + + assert get_llm_provider( + model="iteracompute/qwen/qwen3.8-27b", + custom_llm_provider=None, + api_base="https://explicit.example/v1" if explicit else None, + api_key="explicit-key" if explicit else None, + ) == ( + "qwen/qwen3.8-27b", + "iteracompute", + "explicit-key" if explicit else "env-key", + "https://explicit.example/v1" if explicit else "https://env.example/v1", + ) + + +def test_completion_sends_model_and_bearer_key_without_provider_prefix() -> None: + with respx.mock() as upstream: + route: Final = upstream.post("https://api.iteracompute.com/v1/chat/completions").respond(200, json=_COMPLETION) + response: Final = litellm.completion( + model="iteracompute/qwen/qwen3.8-27b", + messages=[{"role": "user", "content": "Hello"}], + api_key="test-key", + max_tokens=512, + ) + + request: Final = route.calls.last.request + assert request.headers["authorization"] == "Bearer test-key" + assert json.loads(request.content) == { + "model": "qwen/qwen3.8-27b", + "messages": [{"role": "user", "content": "Hello"}], + "max_tokens": 512, + } + assert response.choices[0].message.content == "Hello" + assert response.usage.total_tokens == 7 + + +def test_streaming_returns_text_and_usage() -> None: + event: Final = { + "id": "chatcmpl_test", + "object": "chat.completion.chunk", + "created": 1_790_000_000, + "model": "qwen/qwen3.8-27b", + "choices": [ + { + "index": 0, + "delta": {"role": "assistant", "content": "Hello"}, + "finish_reason": "stop", + } + ], + "usage": {"prompt_tokens": 4, "completion_tokens": 3, "total_tokens": 7}, + } + with respx.mock() as upstream: + route: Final = upstream.post("https://api.iteracompute.com/v1/chat/completions").respond( + 200, + text=f"data: {json.dumps(event)}\n\ndata: [DONE]\n\n", + headers={"content-type": "text/event-stream"}, + ) + chunks: Final = tuple( + litellm.completion( + model="iteracompute/qwen/qwen3.8-27b", + messages=[{"role": "user", "content": "Hello"}], + api_key="test-key", + stream=True, + stream_options={"include_usage": True}, + ) + ) + + body: Final = json.loads(route.calls.last.request.content) + assert body["model"] == "qwen/qwen3.8-27b" + assert body["stream_options"] == {"include_usage": True} + assert "".join(chunk.choices[0].delta.content or "" for chunk in chunks if chunk.choices) == "Hello" + assert next(chunk.usage for chunk in chunks if getattr(chunk, "usage", None) is not None).total_tokens == 7 + + +def test_invalid_key_surfaces_authentication_error() -> None: + with respx.mock() as upstream: + upstream.post("https://api.iteracompute.com/v1/chat/completions").respond( + 401, + json={"error": {"message": "Invalid API key", "type": "authentication_error"}}, + ) + with pytest.raises(litellm.AuthenticationError, match="Invalid API key"): + litellm.completion( + model="iteracompute/qwen/qwen3.8-27b", + messages=[{"role": "user", "content": "Hello"}], + api_key="invalid-test-key", + ) + + +def test_responses_bridge_keeps_model_namespace() -> None: + with respx.mock() as upstream: + route: Final = upstream.post("https://api.iteracompute.com/v1/chat/completions").respond(200, json=_COMPLETION) + response: Final = litellm.responses(model="iteracompute/qwen/qwen3.8-27b", input="Hello", api_key="test-key") + + assert json.loads(route.calls.last.request.content)["model"] == "qwen/qwen3.8-27b" + assert response.output[0].content[0].text == "Hello" + + +@pytest.mark.asyncio +async def test_messages_bridge_keeps_model_namespace( + monkeypatch: pytest.MonkeyPatch, +) -> None: + monkeypatch.setattr(litellm, "disable_aiohttp_transport", True) + monkeypatch.setattr(litellm, "in_memory_llm_clients_cache", LLMClientCache()) + with respx.mock() as upstream: + route: Final = upstream.post("https://api.iteracompute.com/v1/chat/completions").respond(200, json=_COMPLETION) + response: Final = await litellm.anthropic.messages.acreate( + model="iteracompute/qwen/qwen3.8-27b", + messages=[{"role": "user", "content": "Hello"}], + max_tokens=512, + api_key="test-key", + ) + + assert json.loads(route.calls.last.request.content)["model"] == "qwen/qwen3.8-27b" + assert response["content"][0]["text"] == "Hello"