llm-simple-router 0.11.30 → 0.12.0-dev.2464ab6

package/config/recommended-providers.json

@@ -24,7 +24,8 @@
           "deepseek-v4-pro"
         ]
       }
-    ]
+    ],
+    "shortname": "deepseek"
   },
   {
     "group": "百度千帆",
@@ -47,7 +48,8 @@
         ],
         "upstreamPath": "/chat/completions"
       }
-    ]
+    ],
+    "shortname": "qianfan"
   },
   {
     "group": "科大讯飞",
@@ -67,7 +69,8 @@
           "lite"
         ]
       }
-    ]
+    ],
+    "shortname": "iflytek"
   },
   {
     "group": "硅基流动",
@@ -88,7 +91,8 @@
           "moonshotai/Kimi-K2.5"
         ]
       }
-    ]
+    ],
+    "shortname": "siliconflow"
   },
   {
     "group": "智谱",
@@ -139,7 +143,8 @@
           "glm-4.6"
         ]
       }
-    ]
+    ],
+    "shortname": "zhipu"
   },
   {
     "group": "月之暗面",
@@ -169,7 +174,8 @@
           "moonshot-v1-128k"
         ]
       }
-    ]
+    ],
+    "shortname": "moonshot"
   },
   {
     "group": "Minimax",
@@ -199,7 +205,8 @@
           "MiniMax-M2"
         ]
       }
-    ]
+    ],
+    "shortname": "minimax"
   },
   {
     "group": "火山引擎",
@@ -230,7 +237,8 @@
           "doubao-seed-code-preview-251028"
         ]
       }
-    ]
+    ],
+    "shortname": "volcengine"
   },
   {
     "group": "阿里云",
@@ -264,7 +272,8 @@
           "qwen3-coder-next"
         ]
       }
-    ]
+    ],
+    "shortname": "aliyun"
   },
   {
     "group": "腾讯云",
@@ -299,7 +308,8 @@
           "hunyuan-turbos-latest"
         ]
       }
-    ]
+    ],
+    "shortname": "tencent"
   },
   {
     "group": "OpenCode",
@@ -335,7 +345,8 @@
           "minimax-m2.5"
         ]
       }
-    ]
+    ],
+    "shortname": "opencode"
   },
   {
     "group": "阶跃星辰",
@@ -366,6 +377,7 @@
           "step-1-32k"
         ]
       }
-    ]
+    ],
+    "shortname": "stepfun"
   }
 ]

package/config/recommended-retry-rules.json

@@ -1,13 +1,130 @@
 [
-  { "name": "429 Too Many Requests", "status_code": 429, "body_pattern": ".*", "retry_strategy": "exponential", "retry_delay_ms": 5000, "max_retries": 10, "max_delay_ms": 60000, "providers": [] },
-  { "name": "503 Service Unavailable", "status_code": 503, "body_pattern": ".*", "retry_strategy": "exponential", "retry_delay_ms": 5000, "max_retries": 10, "max_delay_ms": 60000, "providers": [] },
-  { "name": "ZAI 网络错误 (code 1234)", "status_code": 400, "body_pattern": "\"type\"\\s*:\\s*\"error\".*\"code\"\\s*:\\s*\"1234\"", "retry_strategy": "exponential", "retry_delay_ms": 5000, "max_retries": 10, "max_delay_ms": 60000, "providers": ["智谱"] },
-  { "name": "ZAI 临时不可用", "status_code": 400, "body_pattern": "\"type\"\\s*:\\s*\"error\".*请稍后重试", "retry_strategy": "exponential", "retry_delay_ms": 5000, "max_retries": 10, "max_delay_ms": 60000, "providers": ["智谱"] },
-  { "name": "ZAI 操作失败 (code 500)", "status_code": 400, "body_pattern": "\"type\"\\s*:\\s*\"error\".*\"code\"\\s*:\\s*\"500\"", "retry_strategy": "exponential", "retry_delay_ms": 5000, "max_retries": 10, "max_delay_ms": 60000, "providers": ["智谱"] },
-  { "name": "ZAI 速率限制 (HTTP 200, code 1302)", "status_code": 200, "body_pattern": "\"error\".*\"code\"\\s*:\\s*\"1302\"", "retry_strategy": "exponential", "retry_delay_ms": 5000, "max_retries": 10, "max_delay_ms": 60000, "providers": ["智谱"] },
-  { "name": "ZAI SSE 错误 (HTTP 200, code 500)", "status_code": 200, "body_pattern": "\"error\".*\"code\"\\s*:\\s*\"500\"", "retry_strategy": "exponential", "retry_delay_ms": 5000, "max_retries": 10, "max_delay_ms": 60000, "providers": ["智谱"] },
-  { "name": "ZAI SSE 错误 (HTTP 200, code 1234)", "status_code": 200, "body_pattern": "\"error\".*\"code\"\\s*:\\s*\"1234\"", "retry_strategy": "exponential", "retry_delay_ms": 5000, "max_retries": 10, "max_delay_ms": 60000, "providers": ["智谱"] },
-  { "name": "ZAI 模型过载 (HTTP 200, code 1305)", "status_code": 200, "body_pattern": "\"error\".*\"code\"\\s*:\\s*\"1305\"", "retry_strategy": "exponential", "retry_delay_ms": 5000, "max_retries": 10, "max_delay_ms": 60000, "providers": ["智谱"] },
-  { "name": "KIMI 401 认证错误", "status_code": 401, "body_pattern": ".*authentication_error.*", "retry_strategy": "exponential", "retry_delay_ms": 5000, "max_retries": 3, "max_delay_ms": 60000, "providers": ["月之暗面"] },
-  { "name": "DeepSeek 并发限流 (429)", "status_code": 429, "body_pattern": "Too many requests.*concurrency", "retry_strategy": "exponential", "retry_delay_ms": 2000, "max_retries": 5, "max_delay_ms": 120000, "providers": ["DeepSeek", "OpenCode"] }
+  {
+    "name": "429 Too Many Requests",
+    "status_code": 429,
+    "body_pattern": ".*",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 5000,
+    "max_retries": 10,
+    "max_delay_ms": 60000,
+    "providers": []
+  },
+  {
+    "name": "503 Service Unavailable",
+    "status_code": 503,
+    "body_pattern": ".*",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 5000,
+    "max_retries": 10,
+    "max_delay_ms": 60000,
+    "providers": []
+  },
+  {
+    "name": "ZAI 网络错误 (code 1234)",
+    "status_code": 400,
+    "body_pattern": "\"type\"\\s*:\\s*\"error\".*\"code\"\\s*:\\s*\"1234\"",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 5000,
+    "max_retries": 10,
+    "max_delay_ms": 60000,
+    "providers": [
+      "zhipu"
+    ]
+  },
+  {
+    "name": "ZAI 临时不可用",
+    "status_code": 400,
+    "body_pattern": "\"type\"\\s*:\\s*\"error\".*请稍后重试",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 5000,
+    "max_retries": 10,
+    "max_delay_ms": 60000,
+    "providers": [
+      "zhipu"
+    ]
+  },
+  {
+    "name": "ZAI 操作失败 (code 500)",
+    "status_code": 400,
+    "body_pattern": "\"type\"\\s*:\\s*\"error\".*\"code\"\\s*:\\s*\"500\"",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 5000,
+    "max_retries": 10,
+    "max_delay_ms": 60000,
+    "providers": [
+      "zhipu"
+    ]
+  },
+  {
+    "name": "ZAI 速率限制 (HTTP 200, code 1302)",
+    "status_code": 200,
+    "body_pattern": "\"error\".*\"code\"\\s*:\\s*\"1302\"",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 5000,
+    "max_retries": 10,
+    "max_delay_ms": 60000,
+    "providers": [
+      "zhipu"
+    ]
+  },
+  {
+    "name": "ZAI SSE 错误 (HTTP 200, code 500)",
+    "status_code": 200,
+    "body_pattern": "\"error\".*\"code\"\\s*:\\s*\"500\"",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 5000,
+    "max_retries": 10,
+    "max_delay_ms": 60000,
+    "providers": [
+      "zhipu"
+    ]
+  },
+  {
+    "name": "ZAI SSE 错误 (HTTP 200, code 1234)",
+    "status_code": 200,
+    "body_pattern": "\"error\".*\"code\"\\s*:\\s*\"1234\"",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 5000,
+    "max_retries": 10,
+    "max_delay_ms": 60000,
+    "providers": [
+      "zhipu"
+    ]
+  },
+  {
+    "name": "ZAI 模型过载 (HTTP 200, code 1305)",
+    "status_code": 200,
+    "body_pattern": "\"error\".*\"code\"\\s*:\\s*\"1305\"",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 5000,
+    "max_retries": 10,
+    "max_delay_ms": 60000,
+    "providers": [
+      "zhipu"
+    ]
+  },
+  {
+    "name": "KIMI 401 认证错误",
+    "status_code": 401,
+    "body_pattern": ".*authentication_error.*",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 5000,
+    "max_retries": 3,
+    "max_delay_ms": 60000,
+    "providers": [
+      "moonshot"
+    ]
+  },
+  {
+    "name": "DeepSeek 并发限流 (429)",
+    "status_code": 429,
+    "body_pattern": "Too many requests.*concurrency",
+    "retry_strategy": "exponential",
+    "retry_delay_ms": 2000,
+    "max_retries": 5,
+    "max_delay_ms": 120000,
+    "providers": [
+      "opencode"
+    ]
+  }
 ]

package/dist/admin/providers.js

@@ -496,7 +496,7 @@ export const adminProviderRoutes = (app, options, done) => {
         base_url: Type.String({ minLength: 1 }),
         models_endpoint: Type.String({ minLength: 1 }),
         api_key: Type.String({ minLength: 1 }),
-        api_type: Type.Union([Type.Literal("openai"), Type.Literal("anthropic")]),
+        api_type: Type.Union([Type.Literal("openai"), Type.Literal("openai-responses"), Type.Literal("anthropic")]),
     });
     app.post("/admin/api/providers/fetch-models", { schema: { body: FetchModelsSchema } }, async (request, reply) => {
         const { base_url, models_endpoint, api_key, api_type } = request.body;

package/dist/admin/quick-setup.js

@@ -1,5 +1,5 @@
 import { Type } from "@sinclair/typebox";
-import { createProvider } from "../db/providers.js";
+import { createProvider, PROVIDER_CONCURRENCY_DEFAULTS } from "../db/providers.js";
 import { createMappingGroup, updateMappingGroup } from "../db/mappings.js";
 import { createRetryRule } from "../db/retry-rules.js";
 import { upsertTransformRule } from "../db/transform-rules.js";
@@ -7,10 +7,36 @@ import { encrypt } from "../utils/crypto.js";
 import { getSetting } from "../db/settings.js";
 import { HTTP_CREATED, HTTP_BAD_REQUEST, HTTP_CONFLICT } from "./constants.js";
 import { API_CODE, apiError } from "./api-response.js";
-import { PROVIDER_CONCURRENCY_DEFAULTS } from "../db/providers.js";
 const PROVIDER_NAME_RE = /^[a-zA-Z0-9_-]+$/;
 const API_KEY_PREVIEW_MIN_LENGTH = 8;
 const API_KEY_PREVIEW_PREFIX_LEN = 4;
+const NEW_PROVIDER_ID = "__new__";
+/** Recursively replace "__new__" provider_id values with the actual provider ID */
+function replaceProviderIds(obj, providerId) {
+    if (Array.isArray(obj)) {
+        return obj.map((item) => replaceProviderIds(item, providerId));
+    }
+    if (obj !== null && typeof obj === "object") {
+        const result = {};
+        for (const [key, value] of Object.entries(obj)) {
+            if ((key === "provider_id" || key === "overflow_provider_id") &&
+                (value === NEW_PROVIDER_ID || value === "")) {
+                result[key] = providerId;
+            }
+            else {
+                result[key] = replaceProviderIds(value, providerId);
+            }
+        }
+        return result;
+    }
+    return obj;
+}
+const QuickSetupEndpointSchema = Type.Object({
+    api_type: Type.Union([Type.Literal("openai"), Type.Literal("openai-responses"), Type.Literal("anthropic")]),
+    base_url: Type.String({ minLength: 1 }),
+    upstream_path: Type.Optional(Type.Union([Type.String({ minLength: 1 }), Type.Null()])),
+    api_key: Type.Optional(Type.Union([Type.String({ minLength: 1 }), Type.Null()])),
+});
 const QuickSetupProviderSchema = Type.Object({
     name: Type.String({ minLength: 1 }),
     api_type: Type.Union([Type.Literal("openai"), Type.Literal("openai-responses"), Type.Literal("anthropic")]),
@@ -21,7 +47,10 @@ const QuickSetupProviderSchema = Type.Object({
         name: Type.String(),
         context_window: Type.Optional(Type.Number()),
         patches: Type.Optional(Type.Array(Type.String())),
+        stream_timeout_ms: Type.Optional(Type.Number()),
+        capabilities: Type.Optional(Type.Array(Type.String())),
     })),
+    endpoints: Type.Optional(Type.Array(QuickSetupEndpointSchema, { minItems: 1 })),
     concurrency_mode: Type.Optional(Type.Union([Type.Literal("auto"), Type.Literal("manual"), Type.Literal("none")])),
     max_concurrency: Type.Optional(Type.Number()),
     queue_timeout_ms: Type.Optional(Type.Number()),
@@ -30,6 +59,9 @@ const QuickSetupProviderSchema = Type.Object({
 const QuickSetupMappingSchema = Type.Object({
     client_model: Type.String({ minLength: 1 }),
     backend_model: Type.String({ minLength: 1 }),
+    /** Optional pre-built rule JSON (targets, overflow, multimodal_fallback).
+     *  If provided, provider_id fields are replaced with the newly created provider's ID. */
+    rule: Type.Optional(Type.String({ minLength: 1 })),
 });
 const QuickSetupRetryRuleSchema = Type.Object({
     name: Type.String({ minLength: 1 }),
@@ -39,6 +71,7 @@ const QuickSetupRetryRuleSchema = Type.Object({
     retry_delay_ms: Type.Number({ minimum: 100 }),
     max_retries: Type.Number({ minimum: 0, maximum: 100 }),
     max_delay_ms: Type.Number({ minimum: 100 }),
+    provider_shortname: Type.Optional(Type.Union([Type.String(), Type.Null()])),
 });
 const QuickSetupTransformSchema = Type.Object({
     inject_headers: Type.Optional(Type.Record(Type.String(), Type.String())),
@@ -82,6 +115,8 @@ export const adminQuickSetupRoutes = (app, options, done) => {
                 name: m.name,
                 ...(m.context_window != null ? { context_window: m.context_window } : {}),
                 ...(m.patches && m.patches.length > 0 ? { patches: m.patches } : {}),
+                ...(m.stream_timeout_ms != null ? { stream_timeout_ms: m.stream_timeout_ms } : {}),
+                ...(m.capabilities && m.capabilities.length > 0 ? { capabilities: m.capabilities } : {}),
             }));
             const adaptiveEnabled = body.provider.concurrency_mode === 'auto' ? 1 : 0;
             const maxConcurrency = body.provider.max_concurrency ?? PROVIDER_CONCURRENCY_DEFAULTS.max_concurrency;
@@ -102,13 +137,31 @@ export const adminQuickSetupRoutes = (app, options, done) => {
                 queue_timeout_ms: queueTimeoutMs,
                 max_queue_size: maxQueueSize,
                 adaptive_enabled: adaptiveEnabled,
+                ...(body.provider.endpoints && body.provider.endpoints.length > 0
+                    ? {
+                        endpoints: JSON.stringify(body.provider.endpoints.map(ep => ({
+                            api_type: ep.api_type,
+                            base_url: ep.base_url,
+                            upstream_path: ep.upstream_path ?? null,
+                            api_key: ep.api_key ? encrypt(ep.api_key, encryptionKey) : (body.provider.api_key ? encrypt(body.provider.api_key, encryptionKey) : null),
+                        }))),
+                    }
+                    : {}),
             });
             // 6. Upsert mapping groups
             for (const m of body.mappings) {
                 const existing = db.prepare('SELECT id FROM mapping_groups WHERE client_model = ?').get(m.client_model);
-                const ruleJson = JSON.stringify({
-                    targets: [{ backend_model: m.backend_model, provider_id: providerId }],
-                });
+                let ruleJson;
+                if (m.rule) {
+                    // Replace provider_id placeholders with the newly created provider's ID
+                    const ruleObj = JSON.parse(m.rule);
+                    ruleJson = JSON.stringify(replaceProviderIds(ruleObj, providerId));
+                }
+                else {
+                    ruleJson = JSON.stringify({
+                        targets: [{ backend_model: m.backend_model, provider_id: providerId }],
+                    });
+                }
                 if (existing) {
                     updateMappingGroup(db, existing.id, {
                         client_model: m.client_model,
@@ -122,8 +175,9 @@ export const adminQuickSetupRoutes = (app, options, done) => {
                     });
                 }
             }
-            // 7. Create retry rules
+            // 7. Create retry rules (bind to newly created provider if shortname matches)
             for (const r of body.retry_rules) {
+                const ruleProviderId = r.provider_shortname ? providerId : null;
                 createRetryRule(db, {
                     name: r.name,
                     status_code: r.status_code,
@@ -133,6 +187,7 @@ export const adminQuickSetupRoutes = (app, options, done) => {
                     retry_delay_ms: r.retry_delay_ms,
                     max_retries: r.max_retries,
                     max_delay_ms: r.max_delay_ms,
+                    provider_id: ruleProviderId,
                 });
             }
             // 8. Create transform rules

package/dist/admin/usage.js

@@ -5,6 +5,8 @@ import { resolveTimeRange } from "../utils/time-range.js";
 const UsageQuerySchema = Type.Object({
     router_key_id: Type.Optional(Type.String()),
     provider_id: Type.Optional(Type.String()),
+    start_time: Type.Optional(Type.String()),
+    end_time: Type.Optional(Type.String()),
 });
 function getDailyUsage(db, startTime, endTime, routerKeyId, providerId) {
     const conditions = [
@@ -41,23 +43,28 @@ export const adminUsageRoutes = (app, options, done) => {
     const { db } = options;
     app.get("/admin/api/usage/windows", { schema: { querystring: UsageQuerySchema } }, async (request) => {
         const query = request.query;
-        if (query.provider_id) {
+        let startTime;
+        let endTime;
+        if (query.start_time && query.end_time) {
+            startTime = query.start_time;
+            endTime = query.end_time;
+        }
+        else if (query.provider_id) {
             const range = resolveTimeRange("window", db, query.router_key_id, query.provider_id);
-            const windows = getWindowsInRange(db, range.startTime, range.endTime, query.router_key_id, query.provider_id);
-            if (windows.length === 0)
-                return [];
-            return windows.map(w => ({
-                window: { ...w, provider_name: resolveProviderName(db, w.provider_id) },
-                usage: getWindowUsage(db, w.start_time, w.end_time, query.router_key_id, query.provider_id),
-            }));
+            startTime = range.startTime;
+            endTime = range.endTime;
+        }
+        else {
+            startTime = "1970-01-01";
+            endTime = "2099-12-31";
         }
-        const allWindows = getWindowsInRange(db, "1970-01-01", "2099-12-31", query.router_key_id)
+        const windows = getWindowsInRange(db, startTime, endTime, query.router_key_id, query.provider_id)
             .filter((w) => w.provider_id !== null);
-        if (allWindows.length === 0)
+        if (windows.length === 0)
             return [];
-        return allWindows.map(w => ({
+        return windows.map(w => ({
             window: { ...w, provider_name: resolveProviderName(db, w.provider_id) },
-            usage: getWindowUsage(db, w.start_time, w.end_time, query.router_key_id),
+            usage: getWindowUsage(db, w.start_time, w.end_time, query.router_key_id, w.provider_id),
         }));
     });
     app.get("/admin/api/usage/weekly", { schema: { querystring: UsageQuerySchema } }, async (request) => {

package/dist/config/recommended-providers.json

@@ -24,7 +24,8 @@
           "deepseek-v4-pro"
         ]
       }
-    ]
+    ],
+    "shortname": "deepseek"
   },
   {
     "group": "百度千帆",
@@ -47,7 +48,8 @@
         ],
         "upstreamPath": "/chat/completions"
       }
-    ]
+    ],
+    "shortname": "qianfan"
   },
   {
     "group": "科大讯飞",
@@ -67,7 +69,8 @@
           "lite"
         ]
       }
-    ]
+    ],
+    "shortname": "iflytek"
   },
   {
     "group": "硅基流动",
@@ -88,7 +91,8 @@
           "moonshotai/Kimi-K2.5"
         ]
       }
-    ]
+    ],
+    "shortname": "siliconflow"
   },
   {
     "group": "智谱",
@@ -139,7 +143,8 @@
           "glm-4.6"
         ]
       }
-    ]
+    ],
+    "shortname": "zhipu"
   },
   {
     "group": "月之暗面",
@@ -169,7 +174,8 @@
           "moonshot-v1-128k"
         ]
       }
-    ]
+    ],
+    "shortname": "moonshot"
   },
   {
     "group": "Minimax",
@@ -199,7 +205,8 @@
           "MiniMax-M2"
         ]
       }
-    ]
+    ],
+    "shortname": "minimax"
   },
   {
     "group": "火山引擎",
@@ -230,7 +237,8 @@
           "doubao-seed-code-preview-251028"
         ]
       }
-    ]
+    ],
+    "shortname": "volcengine"
   },
   {
     "group": "阿里云",
@@ -264,7 +272,8 @@
           "qwen3-coder-next"
         ]
       }
-    ]
+    ],
+    "shortname": "aliyun"
   },
   {
     "group": "腾讯云",
@@ -299,7 +308,8 @@
           "hunyuan-turbos-latest"
         ]
       }
-    ]
+    ],
+    "shortname": "tencent"
   },
   {
     "group": "OpenCode",
@@ -335,7 +345,8 @@
           "minimax-m2.5"
         ]
       }
-    ]
+    ],
+    "shortname": "opencode"
   },
   {
     "group": "阶跃星辰",
@@ -366,6 +377,7 @@
           "step-1-32k"
         ]
       }
-    ]
+    ],
+    "shortname": "stepfun"
   }
 ]