feat: add embeddings endpoint and config for OpenAI-compatible text embeddings via gateway

Add POST /openai/v1/embeddings route in openai_gateway router, new config fields for embedding upstream URL/model/key/enabled toggle with defaults pointing to OpenRouter Qwen3 8B, INTERNAL_EMBED_MODEL constant in config.py, documentation in docs_api.py and README.md describing the molodetz~embed model mapping, and embed-call tracking in gateway metrics alongside existing chat/vision counters.
2026-06-14 01:06:18 +00:00
parent 1076696dec
commit c7770ee21a
17 changed files with 662 additions and 62 deletions
@@ -202,3 +202,85 @@ def test_authorize_require_auth_off_is_open(local_db):
        assert svc.authorize(_make_request_openai_gateway()) is True
    finally:
        set_setting("gateway_require_auth", "1")
+
+
+class FakeEmbedClient_openai_gateway:
+    def __init__(self, *a, **k):
+        self.calls = []
+
+    def build_request(self, method, url, headers=None, json=None, content=None):
+        return FakeRequest(method, url, json)
+
+    async def send(self, request):
+        self.calls.append((request.url, request.json_body))
+        body = request.json_body or {}
+        return FakeResp_openai_gateway(
+            payload={
+                "object": "list",
+                "model": body.get("model"),
+                "data": [{"object": "embedding", "index": 0, "embedding": [0.1, 0.2]}],
+                "usage": {"prompt_tokens": 5, "total_tokens": 5},
+            }
+        )
+
+    async def aclose(self):
+        pass
+
+
+def test_embeddings_remaps_alias_to_configured_model(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeEmbedClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    cfg["gateway_force_model"] = False
+    cfg["gateway_embed_enabled"] = True
+    cfg["gateway_embed_model"] = "qwen/qwen3-embedding-8b"
+    rt = svc.runtime()
+    run_async(
+        rt.handle_embeddings(
+            {"model": "molodetz~embed", "input": "hello"},
+            cfg,
+            ("guest", "test"),
+            "test",
+        )
+    )
+    assert rt._client.calls[-1][1]["model"] == "qwen/qwen3-embedding-8b"
+
+
+def test_embeddings_disabled_returns_503(local_db, monkeypatch):
+    monkeypatch.setattr(gwmod.httpx, "AsyncClient", FakeEmbedClient_openai_gateway)
+    svc = GatewayService()
+    cfg = svc.effective_config()
+    cfg["gateway_embed_enabled"] = False
+    rt = svc.runtime()
+    resp = run_async(
+        rt.handle_embeddings(
+            {"input": "hello"}, cfg, ("guest", "test"), "test"
+        )
+    )
+    assert resp.status_code == 503
+
+
+def test_compute_cost_embed_branch():
+    from devplacepy.services.openai_gateway.usage import (
+        Pricing,
+        compute_cost,
+        normalize_usage,
+    )
+
+    pricing = Pricing(
+        chat_cache_hit_per_m=0.0,
+        chat_cache_miss_per_m=0.0,
+        chat_output_per_m=0.0,
+        vision_input_per_m=0.0,
+        vision_output_per_m=0.0,
+        embed_input_per_m=0.01,
+    )
+    usage = {"prompt_tokens": 1_000_000, "total_tokens": 1_000_000}
+    norm = normalize_usage(usage)
+    total, input_cost, output_cost, native = compute_cost(
+        usage, norm, pricing, "embed"
+    )
+    assert native is False
+    assert output_cost == 0.0
+    assert abs(total - 0.01) < 1e-9
+    assert abs(input_cost - 0.01) < 1e-9