@blockrun/clawrouter 0.12.84 → 0.12.85
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
|
@@ -233,7 +233,7 @@ Edit existing images with `/img2img`:
|
|
|
233
233
|
| openai/gpt-5.2-pro | $21.00 | $168.00 | $0.0945 | 400K | reasoning, tools |
|
|
234
234
|
| openai/gpt-5.4-pro | $30.00 | $180.00 | $0.1050 | 400K | reasoning, tools |
|
|
235
235
|
|
|
236
|
-
> **Free tier:** 11
|
|
236
|
+
> **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
237
|
> **Best value:** `gpt-5-nano` and `gemini-2.5-flash-lite` deliver strong results at ~$0.0003/request.
|
|
238
238
|
|
|
239
239
|
---
|
|
@@ -459,7 +459,7 @@ ClawRouter is an open-source (MIT licensed) smart LLM router built for autonomou
|
|
|
459
459
|
|
|
460
460
|
### How much can ClawRouter save on LLM costs?
|
|
461
461
|
|
|
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
|
|
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 free models ($0/request), while complex tasks get premium models.
|
|
463
463
|
|
|
464
464
|
### How does ClawRouter compare to OpenRouter?
|
|
465
465
|
|
|
@@ -273,4 +273,4 @@ Go build something.
|
|
|
273
273
|
|
|
274
274
|
---
|
|
275
275
|
|
|
276
|
-
*Based on ClawRouter v0.12.
|
|
276
|
+
*Based on ClawRouter v0.12.84. Model availability may change with future releases. For the latest information, visit [blockrun.ai](https://blockrun.ai).*
|
|
@@ -56,7 +56,7 @@ This is where you're paying for real value:
|
|
|
56
56
|
┌─────────────┐ ┌──────────────────────────────┐ ┌──────────────────┐
|
|
57
57
|
│ Your App │────▶│ ClawRouter │────▶│ 41+ AI Models │
|
|
58
58
|
│ (OpenAI │ │ (local proxy) │ │ │
|
|
59
|
-
│ SDK) │ │ │ │ FREE (
|
|
59
|
+
│ SDK) │ │ │ │ FREE (11 free) │
|
|
60
60
|
│ │ │ 1. Route to cheapest model │ │ $0.10 (gemini) │
|
|
61
61
|
│ model: │ │ 2. Compress tokens │ │ $3.00 (sonnet) │
|
|
62
62
|
│ "auto" │ │ 3. Cache repeated requests │ │ $0.20 (grok) │
|
|
@@ -317,7 +317,7 @@ $ /stats 7
|
|
|
317
317
|
| **Request dedup** | Retries don't double-bill | None |
|
|
318
318
|
| **Routing latency** | <1ms (local, on your machine) | Additional network hop |
|
|
319
319
|
| **Payments** | Non-custodial USDC on Base (your wallet, your keys) | Prepaid credit balance (custodial) |
|
|
320
|
-
| **Free tier** |
|
|
320
|
+
| **Free tier** | GPT-OSS-120B (always available) | No free models |
|
|
321
321
|
| **API keys** | Zero — proxy handles all auth | You manage keys per provider |
|
|
322
322
|
| **Algorithm** | Open-source, MIT license, modify it yourself | Proprietary |
|
|
323
323
|
|
|
@@ -147,7 +147,7 @@ x-clawrouter-savings: 82%
|
|
|
147
147
|
x-clawrouter-model: google/gemini-2.5-flash
|
|
148
148
|
```
|
|
149
149
|
|
|
150
|
-
**Per-request USDC payments.** No prepaid balance to drain. Each request shows its price before you pay. When the wallet is empty, requests don't fail — they fall back to the free tier (
|
|
150
|
+
**Per-request USDC payments.** No prepaid balance to drain. Each request shows its price before you pay. When the wallet is empty, requests don't fail — they fall back to the free tier (GPT-OSS-120B).
|
|
151
151
|
|
|
152
152
|
**Budget guard.** `maxCostPerRun` caps per-session spending. Two modes: `graceful` (downgrade to cheaper models) or `strict` (hard stop). The $248/day heartbeat scenario is structurally impossible.
|
|
153
153
|
|
|
@@ -194,7 +194,7 @@ x-clawrouter-reasoning: "Code task with moderate complexity"
|
|
|
194
194
|
| `auto` | Balanced quality + cost | 74–100% |
|
|
195
195
|
| `eco` | Cheapest possible | 95–100% |
|
|
196
196
|
| `premium` | Best quality always | 0% |
|
|
197
|
-
| `free` |
|
|
197
|
+
| `free` | Free models only | 100% |
|
|
198
198
|
|
|
199
199
|
No black box. No mystery routing. Full visibility, full control.
|
|
200
200
|
|
|
@@ -240,7 +240,7 @@ When new models launch, OpenRouter's catalog lags. Users configure a model that
|
|
|
240
240
|
|
|
241
241
|
### How ClawRouter Solves This
|
|
242
242
|
|
|
243
|
-
ClawRouter maintains a curated catalog of 55+ models across 9 providers (including 11 free
|
|
243
|
+
ClawRouter maintains a curated catalog of 55+ models across 9 providers (including 11 free models), updated with each release. Delisted models have automatic redirect aliases:
|
|
244
244
|
|
|
245
245
|
```typescript
|
|
246
246
|
// Delisted models redirect automatically
|
|
@@ -28,7 +28,7 @@ Before building the router, we needed ground truth. We benchmarked all 55+ model
|
|
|
28
28
|
```
|
|
29
29
|
Setup: ClawRouter v0.12.47 proxy on localhost
|
|
30
30
|
→ BlockRun x402 gateway (Base EVM chain)
|
|
31
|
-
→ Provider APIs (OpenAI, Anthropic, Google, xAI, DeepSeek, Moonshot, MiniMax,
|
|
31
|
+
→ Provider APIs (OpenAI, Anthropic, Google, xAI, DeepSeek, Moonshot, MiniMax, Z.AI)
|
|
32
32
|
|
|
33
33
|
Prompts: 3 Python coding tasks (IPv4 validation, LCS algorithm, LRU cache)
|
|
34
34
|
2 requests per model per prompt
|
package/package.json
CHANGED