@botbotgo/better-call 0.1.2 → 0.1.3

package/README.md

@@ -4,7 +4,7 @@
 
 # BetterCall
 
-**One-line wrapper. Three full BFCL remote runs completed. Best: 55.5% → 63.6%.**
+**One-line wrapper. Four full BFCL remote runs completed. Best: 55.5% → 63.6%.**
 
 ```ts
 const tools = betterTools([searchTool, calculatorTool]);
@@ -125,26 +125,27 @@ Latest completed remote run artifact: `benchmarks/bfcl-real-remote-completed-sum
 | --- | ---: | ---: | ---: | ---: | ---: |
 | `qwen3:0.6b` | 3,625 | 55.5% | 63.6% | +8.2pp | 217 |
 | `qwen3.5:0.8b` | 3,625 | 54.6% | 56.9% | +2.3pp | 901 |
+| `qwen3.5:2b` | 3,625 | 53.9% | 54.9% | +1.0pp | 1,308 |
 | `lfm2.5-thinking:latest` | 3,625 | 50.8% | 54.8% | +4.0pp | 1,142 |
 
-Latest completed model category detail: `lfm2.5-thinking:latest`.
+Latest completed model category detail: `qwen3.5:2b`.
 
 | Category | Cases | Raw | BetterCall repair | Lift | Request errors |
 | --- | ---: | ---: | ---: | ---: | ---: |
-| `simple_python` | 400 | 72.5% | 72.5% | +0.0pp | 69 |
-| `simple_java` | 100 | 54.0% | 53.0% | -1.0pp | 11 |
-| `simple_javascript` | 50 | 60.0% | 58.0% | -2.0pp | 4 |
-| `multiple` | 200 | 59.5% | 59.0% | -0.5pp | 65 |
-| `parallel` | 200 | 57.0% | 57.0% | +0.0pp | 61 |
-| `parallel_multiple` | 200 | 26.5% | 26.5% | +0.0pp | 131 |
-| `irrelevance` | 240 | 29.2% | 36.7% | +7.5pp | 152 |
-| `live_simple` | 258 | 57.0% | 57.0% | +0.0pp | 52 |
-| `live_multiple` | 1,053 | 45.9% | 45.9% | +0.0pp | 302 |
-| `live_parallel` | 16 | 43.8% | 43.8% | +0.0pp | 2 |
-| `live_parallel_multiple` | 24 | 45.8% | 45.8% | +0.0pp | 2 |
-| `live_irrelevance` | 884 | 52.3% | 67.1% | +14.8pp | 291 |
-
-The strongest `lfm2.5-thinking:latest` category was `live_irrelevance`: 52.3% raw to 67.1% after BetterCall repair, a +14.8pp lift over 884 real remote cases.
+| `simple_python` | 400 | 29.5% | 32.5% | +3.0pp | 256 |
+| `simple_java` | 100 | 14.0% | 17.0% | +3.0pp | 81 |
+| `simple_javascript` | 50 | 2.0% | 2.0% | +0.0pp | 48 |
+| `multiple` | 200 | 23.0% | 24.0% | +1.0pp | 151 |
+| `parallel` | 200 | 18.0% | 18.0% | +0.0pp | 157 |
+| `parallel_multiple` | 200 | 17.5% | 17.5% | +0.0pp | 159 |
+| `irrelevance` | 240 | 69.6% | 72.5% | +2.9pp | 66 |
+| `live_simple` | 258 | 60.1% | 62.4% | +2.3pp | 41 |
+| `live_multiple` | 1,053 | 68.4% | 68.5% | +0.1pp | 110 |
+| `live_parallel` | 16 | 62.5% | 62.5% | +0.0pp | 2 |
+| `live_parallel_multiple` | 24 | 50.0% | 45.8% | -4.2pp | 6 |
+| `live_irrelevance` | 884 | 72.4% | 73.1% | +0.7pp | 231 |
+
+The strongest `qwen3.5:2b` categories were `simple_python` and `simple_java`, both improving by +3.0pp after BetterCall repair. `live_parallel_multiple` regressed by -4.2pp and is tracked in the full artifact.
 
 Historical targeted wrapper benchmark:
 

package/benchmarks/bfcl-real-remote-completed-summary.json

@@ -2,7 +2,7 @@
   "name": "BFCL v4 full remote Ollama runs",
   "source": "Real model calls against all supported BFCL v4 single-turn tool-call categories completed in this repository",
   "note": "Endpoint redacted. Scores count request errors/timeouts as incorrect. This is not an official BFCL leaderboard submission.",
-  "generatedAt": "2026-05-08T04:26:26.419Z",
+  "generatedAt": "2026-05-08T08:12:17.258Z",
   "results": [
     {
       "model": "qwen3:0.6b",
@@ -240,6 +240,124 @@
         }
       ]
     },
+    {
+      "model": "qwen3.5:2b",
+      "total": 3625,
+      "rawCorrect": 1954,
+      "betterCorrect": 1990,
+      "errors": 1308,
+      "repaired": 55,
+      "categories": [
+        {
+          "model": "qwen3.5:2b",
+          "category": "simple_python",
+          "total": 400,
+          "rawCorrect": 118,
+          "betterCorrect": 130,
+          "errors": 256,
+          "repaired": 15
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "simple_java",
+          "total": 100,
+          "rawCorrect": 14,
+          "betterCorrect": 17,
+          "errors": 81,
+          "repaired": 6
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "simple_javascript",
+          "total": 50,
+          "rawCorrect": 1,
+          "betterCorrect": 1,
+          "errors": 48,
+          "repaired": 0
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "multiple",
+          "total": 200,
+          "rawCorrect": 46,
+          "betterCorrect": 48,
+          "errors": 151,
+          "repaired": 2
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "parallel",
+          "total": 200,
+          "rawCorrect": 36,
+          "betterCorrect": 36,
+          "errors": 157,
+          "repaired": 1
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "parallel_multiple",
+          "total": 200,
+          "rawCorrect": 35,
+          "betterCorrect": 35,
+          "errors": 159,
+          "repaired": 0
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "irrelevance",
+          "total": 240,
+          "rawCorrect": 167,
+          "betterCorrect": 174,
+          "errors": 66,
+          "repaired": 7
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "live_simple",
+          "total": 258,
+          "rawCorrect": 155,
+          "betterCorrect": 161,
+          "errors": 41,
+          "repaired": 8
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "live_multiple",
+          "total": 1053,
+          "rawCorrect": 720,
+          "betterCorrect": 721,
+          "errors": 110,
+          "repaired": 10
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "live_parallel",
+          "total": 16,
+          "rawCorrect": 10,
+          "betterCorrect": 10,
+          "errors": 2,
+          "repaired": 0
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "live_parallel_multiple",
+          "total": 24,
+          "rawCorrect": 12,
+          "betterCorrect": 11,
+          "errors": 6,
+          "repaired": 0
+        },
+        {
+          "model": "qwen3.5:2b",
+          "category": "live_irrelevance",
+          "total": 884,
+          "rawCorrect": 640,
+          "betterCorrect": 646,
+          "errors": 231,
+          "repaired": 6
+        }
+      ]
+    },
     {
       "model": "lfm2.5-thinking:latest",
       "total": 3625,

package/docs/banner.svg

@@ -1,6 +1,6 @@
 <svg xmlns="http://www.w3.org/2000/svg" width="1400" height="520" viewBox="0 0 1400 520" role="img" aria-labelledby="title desc">
   <title id="title">BetterCall banner</title>
-  <desc id="desc">BetterCall is a one-line wrapper with three full BFCL remote runs completed; the best run improves from 55.5 percent to 63.6 percent.</desc>
+  <desc id="desc">BetterCall is a one-line wrapper with four full BFCL remote runs completed; the best run improves from 55.5 percent to 63.6 percent.</desc>
   <defs>
     <linearGradient id="background" x1="0" y1="0" x2="1" y2="1">
       <stop offset="0" stop-color="#07111f"/>
@@ -24,6 +24,6 @@
   <g transform="translate(112 112)">
     <text x="0" y="80" font-family="Inter, ui-sans-serif, system-ui, -apple-system, BlinkMacSystemFont, Segoe UI, sans-serif" font-size="78" font-weight="800" fill="#f8fbff">BetterCall</text>
     <rect x="4" y="116" width="430" height="8" rx="4" fill="url(#accent)"/>
-    <text x="0" y="186" font-family="Inter, ui-sans-serif, system-ui, -apple-system, BlinkMacSystemFont, Segoe UI, sans-serif" font-size="34" font-weight="670" fill="#dcecff">Three full BFCL remote runs. Best: 55.5% → 63.6%.</text>
+    <text x="0" y="186" font-family="Inter, ui-sans-serif, system-ui, -apple-system, BlinkMacSystemFont, Segoe UI, sans-serif" font-size="34" font-weight="670" fill="#dcecff">Four full BFCL remote runs. Best: 55.5% → 63.6%.</text>
   </g>
 </svg>

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@botbotgo/better-call",
-  "version": "0.1.2",
+  "version": "0.1.3",
   "description": "LLM tool-call reliability layer.",
   "type": "module",
   "license": "Apache-2.0",