@blockrun/clawrouter 0.12.84 → 0.12.86

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/README.md CHANGED
@@ -74,17 +74,67 @@ This is the stack that lets agents operate autonomously: **x402 + USDC + local r
74
74
 
75
75
  ## Quick Start
76
76
 
77
+ ### Option A — OpenClaw Agent
78
+
79
+ [OpenClaw](https://openclaw.ai) is an AI coding agent. If you're using it, ClawRouter installs as a plugin:
80
+
77
81
  ```bash
78
- # 1. Install with smart routing enabled
79
82
  curl -fsSL https://blockrun.ai/ClawRouter-update | bash
80
83
  openclaw gateway restart
81
-
82
- # 2. Fund your wallet with USDC on Base or Solana (address printed on install)
83
- # $5 is enough for thousands of requests
84
84
  ```
85
85
 
86
86
  Done. Smart routing (`blockrun/auto`) is now your default model.
87
87
 
88
+ ### Option B — Standalone (continue.dev, Cursor, VS Code, any OpenAI-compatible client)
89
+
90
+ > **Using Claude Code?** Check out [BRCC](https://github.com/BlockRunAI/brcc) — it's purpose-built for Claude Code with the same smart routing and x402 payments.
91
+
92
+ No OpenClaw required. ClawRouter runs as a local proxy on port 8402.
93
+
94
+ **1. Start the proxy**
95
+ ```bash
96
+ npx @blockrun/clawrouter
97
+ ```
98
+
99
+ **2. Fund your wallet**
100
+ Your wallet address is printed on first run. Send a few USDC on Base or Solana — $5 covers thousands of requests.
101
+
102
+ **3. Point your client at `http://localhost:8402`**
103
+
104
+ <details>
105
+ <summary><strong>continue.dev</strong> — <code>~/.continue/config.json</code></summary>
106
+
107
+ ```json
108
+ {
109
+ "models": [
110
+ {
111
+ "title": "ClawRouter Auto",
112
+ "provider": "openai",
113
+ "model": "blockrun/auto",
114
+ "apiBase": "http://localhost:8402",
115
+ "apiKey": "x402"
116
+ }
117
+ ]
118
+ }
119
+ ```
120
+ </details>
121
+
122
+ <details>
123
+ <summary><strong>Cursor</strong> — Settings → Models → OpenAI-compatible</summary>
124
+
125
+ Set base URL to `http://localhost:8402`, API key to `x402`, model to `blockrun/auto`.
126
+ </details>
127
+
128
+ <details>
129
+ <summary><strong>Any OpenAI SDK</strong></summary>
130
+
131
+ ```python
132
+ from openai import OpenAI
133
+ client = OpenAI(base_url="http://localhost:8402", api_key="x402")
134
+ response = client.chat.completions.create(model="blockrun/auto", messages=[...])
135
+ ```
136
+ </details>
137
+
88
138
  ---
89
139
 
90
140
  ## Routing Profiles
@@ -233,7 +283,7 @@ Edit existing images with `/img2img`:
233
283
  | openai/gpt-5.2-pro | $21.00 | $168.00 | $0.0945 | 400K | reasoning, tools |
234
284
  | openai/gpt-5.4-pro | $30.00 | $180.00 | $0.1050 | 400K | reasoning, tools |
235
285
 
236
- > **Free tier:** 11 NVIDIA models cost nothing — `/model free` points to nemotron-ultra-253b, or pick any free model directly (e.g., `/model nemotron`, `/model deepseek-free`, `/model devstral`).
286
+ > **Free tier:** 11 models cost nothing — `/model free` points to nemotron-ultra-253b, or pick any free model directly (e.g., `/model nemotron`, `/model deepseek-free`, `/model devstral`).
237
287
  > **Best value:** `gpt-5-nano` and `gemini-2.5-flash-lite` deliver strong results at ~$0.0003/request.
238
288
 
239
289
  ---
@@ -459,7 +509,7 @@ ClawRouter is an open-source (MIT licensed) smart LLM router built for autonomou
459
509
 
460
510
  ### How much can ClawRouter save on LLM costs?
461
511
 
462
- ClawRouter's blended average cost is $2.05 per million tokens compared to $25/M for Claude Opus, representing 92% savings. Actual savings depend on your workload — simple queries are routed to budget models (as low as $0/request with NVIDIA free tier), while complex tasks get premium models.
512
+ ClawRouter's blended average cost is $2.05 per million tokens compared to $25/M for Claude Opus, representing 92% savings. Actual savings depend on your workload — simple queries are routed to free models ($0/request), while complex tasks get premium models.
463
513
 
464
514
  ### How does ClawRouter compare to OpenRouter?
465
515
 
@@ -471,7 +521,7 @@ Both are open source and run locally. But ClawRouter adds smart routing (automat
471
521
 
472
522
  ### What agents does ClawRouter work with?
473
523
 
474
- ClawRouter integrates with OpenClaw (Claude Code), ElizaOS, and any agent that makes OpenAI-compatible API calls. It runs as a local proxy on port 8402.
524
+ ClawRouter works with any tool that makes OpenAI-compatible API calls — point it at `http://localhost:8402`. This includes continue.dev, Cursor, VS Code extensions, ElizaOS, and custom agents. It also integrates as a plugin with [OpenClaw](https://openclaw.ai) (an AI coding agent), which enables additional features like slash commands and usage reports.
475
525
 
476
526
  ### Is ClawRouter free?
477
527
 
package/dist/cli.js CHANGED
@@ -49885,9 +49885,9 @@ data: [DONE]
49885
49885
  maxTokens,
49886
49886
  routingProfile ?? void 0
49887
49887
  );
49888
- logCost = costs.costEstimate;
49888
+ logCost = FREE_MODELS.has(logModel) ? 0 : costs.costEstimate;
49889
49889
  logBaseline = costs.baselineCost;
49890
- logSavings = costs.savings;
49890
+ logSavings = FREE_MODELS.has(logModel) ? 1 : costs.savings;
49891
49891
  }
49892
49892
  const entry = {
49893
49893
  timestamp: (/* @__PURE__ */ new Date()).toISOString(),