fix(openclaw): fix 404 on inference, add missing models, reduce log spam

- Fix upstream URL: append /v1 to DEFAULT_SURF_UPSTREAM_URL so chat/completions
  routes correctly to surf.cascade.fyi/api/v1/inference/v1/chat/completions
- Add 5 missing models to MODEL_METADATA: z-ai/glm-5.1,
  x-ai/grok-4.20-multi-agent-beta, and 3 :online variants
- Wire fetchUpstreamModels into catalog run for auto-sync with live API
- Downgrade startup logs from info to debug to reduce plugin log spam
- Update reference doc with all 19 models and correct upstream URL

Author: tenequm

packages/x402-proxy/skills/references/openclaw-plugin.md

@@ -41,7 +41,7 @@ If you do nothing, the plugin now defaults to:
 
 - provider id: `surf`
 - gateway base URL: `/x402-proxy/v1`
-- upstream: `https://surf.cascade.fyi/api/v1/inference`
+- upstream: `https://surf.cascade.fyi/api/v1/inference/v1`
 - protocol: `mpp`
 
 You only need explicit config if you want to override the defaults or supply a custom model catalog.
@@ -57,23 +57,27 @@ Add the plugin config to your `openclaw.json` (or via `openclaw config edit`):
           "providers": {
             "surf-inference": {
               "baseUrl": "/x402-proxy/v1",
-              "upstreamUrl": "https://surf.cascade.fyi/api/v1/inference",
+              "upstreamUrl": "https://surf.cascade.fyi/api/v1/inference/v1",
               "models": [
                 { "id": "anthropic/claude-opus-4.6", "name": "Claude Opus 4.6", "maxTokens": 200000, "reasoning": true, "input": ["text", "image"], "cost": { "input": 0.015, "output": 0.075, "cacheRead": 0.0015, "cacheWrite": 0.01875 }, "contextWindow": 200000 },
-                { "id": "anthropic/claude-sonnet-4.6", "name": "Claude Sonnet 4.6", "maxTokens": 200000, "reasoning": true, "input": ["text", "image"], "cost": { "input": 0.003, "output": 0.015, "cacheRead": 0.0003, "cacheWrite": 0.00375 }, "contextWindow": 200000 },
                 { "id": "anthropic/claude-opus-4.5", "name": "Claude Opus 4.5", "maxTokens": 200000, "reasoning": true, "input": ["text", "image"], "cost": { "input": 0.015, "output": 0.075, "cacheRead": 0.0015, "cacheWrite": 0.01875 }, "contextWindow": 200000 },
+                { "id": "anthropic/claude-sonnet-4.6", "name": "Claude Sonnet 4.6", "maxTokens": 200000, "reasoning": true, "input": ["text", "image"], "cost": { "input": 0.003, "output": 0.015, "cacheRead": 0.0003, "cacheWrite": 0.00375 }, "contextWindow": 200000 },
                 { "id": "anthropic/claude-sonnet-4.5", "name": "Claude Sonnet 4.5", "maxTokens": 200000, "reasoning": true, "input": ["text", "image"], "cost": { "input": 0.003, "output": 0.015, "cacheRead": 0.0003, "cacheWrite": 0.00375 }, "contextWindow": 200000 },
                 { "id": "x-ai/grok-4.20-beta", "name": "Grok 4.20 Beta", "maxTokens": 131072, "reasoning": true, "input": ["text"], "cost": { "input": 0.003, "output": 0.015, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
                 { "id": "x-ai/grok-4.20-multi-agent-beta", "name": "Grok 4.20 Multi-Agent", "maxTokens": 131072, "reasoning": true, "input": ["text"], "cost": { "input": 0.003, "output": 0.015, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
                 { "id": "x-ai/grok-4.1-fast", "name": "Grok 4.1 Fast", "maxTokens": 131072, "reasoning": false, "input": ["text"], "cost": { "input": 0.001, "output": 0.005, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
-                { "id": "x-ai/grok-4.20-beta:online", "name": "Grok 4.20 Beta (Online)", "maxTokens": 131072, "reasoning": true, "input": ["text"], "cost": { "input": 0.005, "output": 0.025, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
-                { "id": "x-ai/grok-4.20-multi-agent-beta:online", "name": "Grok 4.20 Multi-Agent (Online)", "maxTokens": 131072, "reasoning": true, "input": ["text"], "cost": { "input": 0.005, "output": 0.025, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
-                { "id": "x-ai/grok-4.1-fast:online", "name": "Grok 4.1 Fast (Online)", "maxTokens": 131072, "reasoning": false, "input": ["text"], "cost": { "input": 0.003, "output": 0.015, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
+                { "id": "x-ai/grok-4.20-beta:online", "name": "Grok 4.20 Beta (Online)", "maxTokens": 131072, "reasoning": true, "input": ["text"], "cost": { "input": 0.004, "output": 0.015, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
+                { "id": "x-ai/grok-4.20-multi-agent-beta:online", "name": "Grok 4.20 Multi-Agent (Online)", "maxTokens": 131072, "reasoning": true, "input": ["text"], "cost": { "input": 0.004, "output": 0.015, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
+                { "id": "x-ai/grok-4.1-fast:online", "name": "Grok 4.1 Fast (Online)", "maxTokens": 131072, "reasoning": false, "input": ["text"], "cost": { "input": 0.001, "output": 0.005, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
                 { "id": "minimax/minimax-m2.7", "name": "MiniMax M2.7", "maxTokens": 1000000, "reasoning": false, "input": ["text"], "cost": { "input": 0.001, "output": 0.005, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 1000000 },
                 { "id": "minimax/minimax-m2.5", "name": "MiniMax M2.5", "maxTokens": 1000000, "reasoning": false, "input": ["text"], "cost": { "input": 0.001, "output": 0.005, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 1000000 },
                 { "id": "moonshotai/kimi-k2.5", "name": "Kimi K2.5", "maxTokens": 131072, "reasoning": true, "input": ["text"], "cost": { "input": 0.002, "output": 0.008, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
                 { "id": "z-ai/glm-5", "name": "GLM-5", "maxTokens": 128000, "reasoning": false, "input": ["text"], "cost": { "input": 0.001, "output": 0.005, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 128000 },
-                { "id": "qwen/qwen-2.5-7b-instruct", "name": "Qwen 2.5 7B Instruct", "maxTokens": 32768, "reasoning": false, "input": ["text"], "cost": { "input": 0.0003, "output": 0.001, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 32768 }
+                { "id": "z-ai/glm-5-turbo", "name": "GLM-5 Turbo", "maxTokens": 128000, "reasoning": false, "input": ["text"], "cost": { "input": 0.001, "output": 0.005, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 128000 },
+                { "id": "z-ai/glm-5.1", "name": "GLM-5.1", "maxTokens": 202000, "reasoning": true, "input": ["text"], "cost": { "input": 0.002, "output": 0.006, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 202000 },
+                { "id": "qwen/qwen-2.5-7b-instruct", "name": "Qwen 2.5 7B Instruct", "maxTokens": 32768, "reasoning": false, "input": ["text"], "cost": { "input": 0.0005, "output": 0.002, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
+                { "id": "stepfun/step-3.5-flash", "name": "Step 3.5 Flash", "maxTokens": 131072, "reasoning": false, "input": ["text"], "cost": { "input": 0.001, "output": 0.005, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 },
+                { "id": "xiaomi/mimo-v2-pro", "name": "MiMo V2 Pro", "maxTokens": 131072, "reasoning": true, "input": ["text"], "cost": { "input": 0.001, "output": 0.005, "cacheRead": 0, "cacheWrite": 0 }, "contextWindow": 131072 }
               ]
             }
           }
@@ -135,10 +139,17 @@ npx x402-proxy --protocol mpp --network solana \
   https://surf.cascade.fyi/api/v1/inference/v1/models
 ```
 
+Or use the catalog auto-sync: the plugin now fetches the live model list from the upstream on each gateway restart. Just restart OpenClaw:
+
+```bash
+openclaw gateway restart
+```
+
 Then update the `models` array in your plugin config accordingly.
 
 ## Troubleshooting
 
+- **HTTP 404 on inference** - check your `upstreamUrl` ends with `/v1` (e.g., `https://surf.cascade.fyi/api/v1/inference/v1`). OpenClaw appends `/chat/completions` directly to the baseUrl, so the `/v1` must be in the upstream path.
 - **Models don't appear in `openclaw models`** - the plugin uses a `catalog` hook (not `models` field). Make sure you're on x402-proxy >= 0.8.5.
 - **"no wallet found" in logs** - run `npx x402-proxy setup` or set `X402_PROXY_WALLET_MNEMONIC` env var before starting the gateway.
 - **402 errors on inference** - check wallet has USDC balance: use `x_wallet` tool or `npx x402-proxy wallet`.

packages/x402-proxy/src/openclaw/defaults.ts

@@ -21,7 +21,7 @@ export type ResolvedProviderConfig = {
 
 export const DEFAULT_SURF_PROVIDER_ID = "surf";
 export const DEFAULT_SURF_BASE_URL = "/x402-proxy/v1";
-export const DEFAULT_SURF_UPSTREAM_URL = "https://surf.cascade.fyi/api/v1/inference";
+export const DEFAULT_SURF_UPSTREAM_URL = "https://surf.cascade.fyi/api/v1/inference/v1";
 export const DEFAULT_PROVIDER_PROTOCOL: PaymentProtocol = "mpp";
 export const DEFAULT_MPP_SESSION_BUDGET = "0.5";
 
@@ -115,6 +115,46 @@ const MODEL_METADATA: Record<string, Omit<ModelEntry, "provider" | "id">> = {
     cost: { input: 0.001, output: 0.005, cacheRead: 0, cacheWrite: 0 },
     contextWindow: 128000,
   },
+  "z-ai/glm-5.1": {
+    name: "GLM-5.1",
+    maxTokens: 202000,
+    reasoning: true,
+    input: ["text"],
+    cost: { input: 0.002, output: 0.006, cacheRead: 0, cacheWrite: 0 },
+    contextWindow: 202000,
+  },
+  "x-ai/grok-4.20-multi-agent-beta": {
+    name: "Grok 4.20 Multi-Agent Beta",
+    maxTokens: 131072,
+    reasoning: true,
+    input: ["text"],
+    cost: { input: 0.003, output: 0.015, cacheRead: 0, cacheWrite: 0 },
+    contextWindow: 131072,
+  },
+  "x-ai/grok-4.1-fast:online": {
+    name: "Grok 4.1 Fast (Online)",
+    maxTokens: 131072,
+    reasoning: false,
+    input: ["text"],
+    cost: { input: 0.001, output: 0.005, cacheRead: 0, cacheWrite: 0 },
+    contextWindow: 131072,
+  },
+  "x-ai/grok-4.20-beta:online": {
+    name: "Grok 4.20 Beta (Online)",
+    maxTokens: 131072,
+    reasoning: true,
+    input: ["text"],
+    cost: { input: 0.004, output: 0.015, cacheRead: 0, cacheWrite: 0 },
+    contextWindow: 131072,
+  },
+  "x-ai/grok-4.20-multi-agent-beta:online": {
+    name: "Grok 4.20 Multi-Agent Beta (Online)",
+    maxTokens: 131072,
+    reasoning: true,
+    input: ["text"],
+    cost: { input: 0.004, output: 0.015, cacheRead: 0, cacheWrite: 0 },
+    contextWindow: 131072,
+  },
   "qwen/qwen-2.5-7b-instruct": {
     name: "Qwen 2.5 7B Instruct",
     maxTokens: 32768,
@@ -174,7 +214,7 @@ export async function fetchUpstreamModels(
     return modelsCache.models;
   }
   try {
-    const res = await globalThis.fetch(`${upstreamUrl}/v1/models`, {
+    const res = await globalThis.fetch(`${upstreamUrl}/models`, {
       signal: AbortSignal.timeout(5000),
     });
     if (!res.ok) return modelsCache?.models ?? DEFAULT_SURF_MODELS;

packages/x402-proxy/src/openclaw/plugin.ts

@@ -19,7 +19,7 @@ import { OptimizedSvmScheme } from "../lib/optimized-svm-scheme.js";
 import { resolveWallet } from "../lib/wallet-resolution.js";
 import { loadSvmWallet } from "../wallet.js";
 import { createSendCommand, createWalletCommand } from "./commands.js";
-import { resolveProviders, routePrefixForBaseUrl } from "./defaults.js";
+import { fetchUpstreamModels, resolveProviders, routePrefixForBaseUrl } from "./defaults.js";
 import { createInferenceProxyRouteHandler } from "./route.js";
 import { addressForNetwork, createRequestTool, createWalletTool, SOL_MAINNET } from "./tools.js";
 
@@ -152,21 +152,24 @@ export function register(api: OpenClawPluginApi): void {
         run: async (ctx) => {
           const { apiKey } = ctx.resolveProviderApiKey(provider.id);
           if (!apiKey) return null;
+          const models = await fetchUpstreamModels(provider.upstreamUrl).catch(
+            () => provider.models,
+          );
           return {
             provider: {
               baseUrl: provider.baseUrl,
               api: "openai-completions",
               authHeader: false,
               apiKey,
-              models: provider.models,
+              models,
             },
           };
         },
       },
     });
   }
 
-  api.logger.info(
+  api.logger.debug?.(
     `x402-proxy: ${providers.map((provider) => `${provider.id}:${provider.protocol}`).join(", ")} - ${allModels.length} models`,
   );
 
@@ -202,7 +205,7 @@ export function register(api: OpenClawPluginApi): void {
       handler,
     });
   }
-  api.logger.info(
+  api.logger.debug?.(
     `x402-proxy: HTTP routes ${routePrefixes.join(", ")} registered for ${providers.map((provider) => provider.upstreamUrl).join(", ")}`,
   );
 
@@ -276,7 +279,7 @@ export function register(api: OpenClawPluginApi): void {
           mppHandlerRef = null;
         }
 
-        api.logger.info(
+        api.logger.debug?.(
           `wallets: solana=${solanaWalletAddress ?? "missing"} evm=${evmWalletAddress ?? "missing"}`,
         );
       } catch (err) {