Release: BYOK web + qwen models + CI cleanup

.github/PULL_REQUEST_TEMPLATE.md

@@ -15,7 +15,6 @@ Brief description of changes.
 
 - [ ] Tests pass locally
 - [ ] PoW solvers tested (C + JS)
-- [ ] Live tests if applicable
 
 ## Checklist
 

.github/workflows/ci.yml

@@ -48,36 +48,6 @@ jobs:
       - name: Run tests
         run: python tests.py -j 2
 
-  live-test:
-    needs: test
-    runs-on: ubuntu-latest
-    timeout-minutes: 30
-    if: github.event_name == 'push'
-    steps:
-      - uses: actions/checkout@v7
-
-      - uses: actions/setup-python@v7
-        with:
-          python-version: "3.12"
-          cache: pip
-          cache-dependency-path: requirements-dev.txt
-
-      - name: Install system tools
-        run: |
-          sudo apt-get update
-          sudo apt-get install -y --no-install-recommends clang nodejs
-
-      - name: Install dependencies
-        run: pip install -r requirements-dev.txt
-
-      - name: Build native PoW solver
-        run: clang -O3 -pthread -funroll-loops -flto -march=native -mtune=native -o danyapi/deepseek/pow_solver danyapi/deepseek/pow_solver.c
-
-      - name: Run live tests
-        env:
-          DEEPSEEK_TOKENS: ${{ secrets.DEEPSEEK_TOKENS }}
-        run: python -m pytest tests/test_live.py -v
-
   docker-build:
     needs: test
     runs-on: ubuntu-latest

danyapi/api/openai.py

@@ -1046,6 +1046,8 @@ async def list_models() -> dict:
                 }
             )
     qwen_models: list[dict] = getattr(app.state, "qwen_models", [])
+    if not qwen_models and _byok_mode():
+        qwen_models = QWEN_DEFAULT_MODELS
     for model in qwen_models:
         models.append(
             {

pyproject.toml

@@ -9,9 +9,6 @@ select = ["B", "C4", "E", "F", "FURB", "I", "PLW", "RUF", "UP"]
 testpaths = ["tests"]
 asyncio_mode = "auto"
 filterwarnings = ["ignore:Using `httpx` with `starlette.testclient` is deprecated.*"]
-markers = [
-    "live: tests that hit the real DeepSeek API (requires DEEPSEEK_TOKENS env var)",
-]
 
 [tool.mypy]
 python_version = "3.12"

tests/test_byok.py

@@ -228,6 +228,18 @@ def test_health_reports_byok_mode():
     assert payload["byok_pools"]["qwen"] == 0
 
 
+def test_list_models_byok_without_env_tokens_lists_qwen_defaults():
+    app.state.byok = True
+    app.state.qwen_models = []
+    client = TestClient(app)
+    data = client.get("/v1/models").json()
+    client.close()
+    ids = [m["id"] for m in data["data"]]
+    assert "deepseek-v4.1-flash" in ids
+    for model in openai_mod.QWEN_DEFAULT_MODELS:
+        assert model["id"] in ids
+
+
 def test_health_no_byok_key_when_disabled():
     app.state.byok = False
     client = TestClient(app)

web/index.html

@@ -530,29 +530,33 @@
             </svg>
             <span class="brand-name">Dany<span class="brand-accent">API</span></span>
         </a>
-        <h1><span class="grad-text">Token Manager</span></h1>
-        <p>Support the project add your free provider tokens</p>
+        <div id="token-heading">
+            <h1><span class="grad-text">Token Manager</span></h1>
+            <p>Support the project add your free provider tokens</p>
+        </div>
     </div>
 
-    <div class="form-group">
-        <label for="deepseek-tokens">DeepSeek Tokens</label>
-        <input type="text" id="deepseek-tokens"
-               placeholder="token1, token2, token3..." value="" autocomplete="off" spellcheck="false" />
-    </div>
+    <section class="token-manager" id="token-manager">
+        <div class="form-group">
+            <label for="deepseek-tokens">DeepSeek Tokens</label>
+            <input type="text" id="deepseek-tokens"
+                   placeholder="token1, token2, token3..." value="" autocomplete="off" spellcheck="false" />
+        </div>
 
-    <div class="form-group">
-        <label for="qwen-tokens">Qwen Tokens</label>
-        <input type="text" id="qwen-tokens"
-               placeholder="token1, token2, token3..." value="" autocomplete="off" spellcheck="false" />
-    </div>
+        <div class="form-group">
+            <label for="qwen-tokens">Qwen Tokens</label>
+            <input type="text" id="qwen-tokens"
+                   placeholder="token1, token2, token3..." value="" autocomplete="off" spellcheck="false" />
+        </div>
 
-    <div class="info-box">
-        <strong>No limits.</strong> Tokens are used server-side. Your consumers just point their OpenAI client at this instance no keys needed on their side. Each token adds parallel generation capacity.
-    </div>
+        <div class="info-box">
+            <strong>No limits.</strong> Tokens are used server-side. Your consumers just point their OpenAI client at this instance no keys needed on their side. Each token adds parallel generation capacity.
+        </div>
 
-    <div id="status" class="status"></div>
+        <div id="status" class="status"></div>
 
-    <button class="btn btn-primary" type="button" id="btn-save">Add Tokens</button>
+        <button class="btn btn-primary" type="button" id="btn-save">Add Tokens</button>
+    </section>
 
     <div class="usage-section">
         <h2 class="usage-title">Usage</h2>
@@ -617,6 +621,24 @@ <h2 class="usage-title">Usage</h2>
             });
     }
 
+    function initMode() {
+        fetch(BASE + "/health")
+            .then(function (r) {
+                if (!r.ok) throw new Error("HTTP " + r.status);
+                return r.json();
+            })
+            .then(function (data) {
+                if (data.byok) {
+                    var heading = document.getElementById("token-heading");
+                    if (heading) heading.style.display = "none";
+                    var manager = document.getElementById("token-manager");
+                    if (manager) manager.style.display = "none";
+                }
+            })
+            .catch(function () {});
+    }
+
+    initMode();
     loadUsage();
 
     function showStatus(msg, type) {