feat(providers): add Standard Compute chat completions

litellm/llms/openai_like/providers.json

@@ -200,5 +200,12 @@
       "temperature_max": 1.99
     },
     "supported_endpoints": ["/v1/chat/completions"]
+  },
+  "standardcompute": {
+    "base_url": "https://api.stdcmpt.com/v1",
+    "api_key_env": "STANDARDCOMPUTE_API_KEY",
+    "param_mappings": {
+      "max_completion_tokens": "max_tokens"
+    }
   }
 }

tests/test_litellm/llms/openai_like/test_standardcompute_provider.py

@@ -0,0 +1,158 @@
+import json
+
+import httpx
+import pytest
+import respx
+
+import litellm
+
+
+@pytest.fixture(autouse=True)
+def httpx_transport(monkeypatch):
+    monkeypatch.setenv("DISABLE_AIOHTTP_TRANSPORT", "True")
+
+
+@pytest.mark.asyncio
+async def test_standardcompute_completion_preserves_tools_and_maps_token_limit():
+    tools = [
+        {
+            "type": "function",
+            "function": {
+                "name": "read_file",
+                "parameters": {
+                    "type": "object",
+                    "properties": {"path": {"type": "string"}},
+                    "required": ["path"],
+                },
+            },
+        }
+    ]
+    response = {
+        "id": "chatcmpl-test",
+        "object": "chat.completion",
+        "created": 1,
+        "model": "standardcompute",
+        "choices": [
+            {
+                "index": 0,
+                "finish_reason": "tool_calls",
+                "message": {
+                    "role": "assistant",
+                    "content": None,
+                    "tool_calls": [
+                        {
+                            "id": "call-read",
+                            "type": "function",
+                            "function": {
+                                "name": "read_file",
+                                "arguments": '{"path":"README.md"}',
+                            },
+                        }
+                    ],
+                },
+            }
+        ],
+    }
+    with respx.mock as transport:
+        request = transport.post("https://api.stdcmpt.com/v1/chat/completions").mock(
+            return_value=httpx.Response(200, json=response)
+        )
+        result = await litellm.acompletion(
+            model="standardcompute/standardcompute",
+            api_key="test-key",
+            messages=[{"role": "user", "content": "Read README.md"}],
+            tools=tools,
+            max_completion_tokens=64,
+        )
+    sent = json.loads(request.calls.last.request.content)
+    assert sent["model"] == "standardcompute"
+    assert sent["tools"] == tools
+    assert sent["max_tokens"] == 64
+    assert "max_completion_tokens" not in sent
+    assert request.calls.last.request.headers["authorization"] == "Bearer test-key"
+    assert result.choices[0].message.tool_calls[0].function.name == "read_file"
+
+
+@pytest.mark.asyncio
+async def test_standardcompute_environment_key_and_api_base_override(monkeypatch):
+    monkeypatch.setenv("STANDARDCOMPUTE_API_KEY", "environment-test-key")
+    response = {
+        "id": "chatcmpl-test",
+        "object": "chat.completion",
+        "created": 1,
+        "model": "standardcompute",
+        "choices": [
+            {
+                "index": 0,
+                "finish_reason": "stop",
+                "message": {"role": "assistant", "content": "Connected"},
+            }
+        ],
+    }
+    with respx.mock as transport:
+        request = transport.post("https://example.test/v1/chat/completions").mock(
+            return_value=httpx.Response(200, json=response)
+        )
+        result = await litellm.acompletion(
+            model="standardcompute/standardcompute",
+            api_base="https://example.test/v1",
+            messages=[{"role": "user", "content": "Connection check"}],
+        )
+    assert (
+        request.calls.last.request.headers["authorization"]
+        == "Bearer environment-test-key"
+    )
+    assert result.choices[0].message.content == "Connected"
+
+
+@pytest.mark.asyncio
+async def test_standardcompute_stream_consumes_chat_completion_events():
+    chunks = [
+        {
+            "id": "chatcmpl-test",
+            "object": "chat.completion.chunk",
+            "created": 1,
+            "model": "standardcompute",
+            "choices": [
+                {
+                    "index": 0,
+                    "delta": {"role": "assistant", "content": "Connected"},
+                    "finish_reason": None,
+                }
+            ],
+        },
+        {
+            "id": "chatcmpl-test",
+            "object": "chat.completion.chunk",
+            "created": 1,
+            "model": "standardcompute",
+            "choices": [{"index": 0, "delta": {}, "finish_reason": "stop"}],
+        },
+    ]
+    events = (
+        "".join("data: " + json.dumps(chunk) + "\n\n" for chunk in chunks)
+        + "data: [DONE]\n\n"
+    )
+    with respx.mock as transport:
+        request = transport.post("https://api.stdcmpt.com/v1/chat/completions").mock(
+            return_value=httpx.Response(
+                200, text=events, headers={"content-type": "text/event-stream"}
+            )
+        )
+        stream = await litellm.acompletion(
+            model="standardcompute/standardcompute",
+            api_key="test-key",
+            messages=[{"role": "user", "content": "Connection check"}],
+            stream=True,
+        )
+        received = [chunk async for chunk in stream]
+    assert json.loads(request.calls.last.request.content)["stream"] is True
+    assert (
+        "".join(
+            chunk.choices[0].delta.content or "" for chunk in received if chunk.choices
+        )
+        == "Connected"
+    )
+    assert any(
+        chunk.choices and chunk.choices[0].finish_reason == "stop" for chunk in received
+    )