# The Multivac — Evaluation Report

**Evaluation ID:** EVAL-20260207-142929
**Date:** Mar 17, 2026
**Category:** code
**Question ID:** CODE-010

---

## Question

Create TypeScript types that enforce these compile-time constraints:

1. A `Route` type where the path parameters in the URL template must match the params object:
   - `createRoute('/users/:userId/posts/:postId', { userId: '123', postId: '456' })` ✓
   - `createRoute('/users/:userId', { wrongParam: '123' })` ✗ (compile error)

2. A `Builder` pattern type where methods can only be called in a specific order:
   - `builder.setName('x').setAge(25).build()` ✓
   - `builder.setAge(25).setName('x').build()` ✗ (compile error)
   - `builder.setName('x').build()` ✗ (compile error - age not set)

Show the type definitions and example usage.

---

## Winner

**Claude Sonnet 4.5** (Anthropic)
- Winner Score: 9.49
- Matrix Average: 7.87
- Total Judgments: 90

---

## Rankings

| Rank | Model | Provider | Avg Score | Judgments |
|------|-------|----------|-----------|----------|
| 1 | Claude Sonnet 4.5 | Anthropic | 9.49 | 8 |
| 2 | Grok 3 (Direct) | xAI | 9.42 | 8 |
| 3 | GPT-5.2-Codex | OpenAI | 9.36 | 8 |
| 4 | Gemini 3 Flash Preview | Google | 9.28 | 8 |
| 5 | Grok Code Fast | xAI | 9.16 | 8 |
| 6 | Claude Opus 4.5 | Anthropic | 9.07 | 8 |
| 7 | DeepSeek V3.2 | DeepSeek | 8.84 | 7 |
| 8 | Gemini 3 Pro Preview | Google | 5.62 | 9 |
| 9 | GLM-4-7 | Zhipu | 4.25 | 4 |
| 10 | MiniMax M2 | MiniMax | 4.24 | 4 |

---

## 10×10 Judgment Matrix

Rows = Judge, Columns = Respondent. Self-judgments excluded (—).

| Judge ↓ / Resp → | Grok Code Fast | GLM-4-7 | Claude Opus | Gemini 3 | Claude Sonnet | Gemini 3 | MiniMax M2 | DeepSeek V3.2 | GPT-5.2-Codex | Grok 3 |
|---|---|---|---|---|---|---|---|---|---|---|
| Grok Code Fast | — | 0.0 | 9.8 | 9.8 | 9.6 | 5.5 | 0.0 | 9.6 | 9.8 | 9.6 |
| GLM-4-7 | 9.8 | — | 9.3 | 9.8 | 9.8 | 5.8 | 0.0 | 0.0 | 9.8 | 9.8 |
| Claude Opus | 8.3 | 0.9 | — | 9.0 | 9.0 | 5.3 | 1.3 | 6.8 | 8.6 | 9.0 |
| Gemini 3 | 9.8 | 0.0 | 9.6 | — | 10.0 | 6.5 | 0.0 | 9.8 | 9.8 | 9.8 |
| Claude Sonnet | 9.2 | 0.5 | 9.4 | 9.3 | — | 4.7 | 0.5 | 8.4 | 8.6 | 9.2 |
| Gemini 3 | 0.0 | 0.0 | 0.0 | 0.0 | 0.0 | — | 0.0 | 0.0 | 0.0 | 0.0 |
| MiniMax M2 | 10.0 | 0.0 | 8.4 | 9.8 | 9.8 | 4.7 | — | 10.0 | 9.4 | 9.8 |
| DeepSeek V3.2 | 8.8 | 8.6 | 8.8 | 8.6 | 9.3 | 7.2 | 8.2 | — | 9.8 | 9.3 |
| GPT-5.2-Codex | 8.8 | 0.0 | 7.8 | 8.8 | 8.8 | 4.1 | 0.0 | 8.6 | — | 8.8 |
| Grok 3 | 8.6 | 7.0 | 9.4 | 9.1 | 9.6 | 6.9 | 7.0 | 8.8 | 9.1 | — |

---

## Methodology

- **10×10 Blind Peer Matrix:** All models answer the same question, then all models judge all responses.
- **5 Criteria:** Correctness, completeness, clarity, depth, usefulness (each scored 1–10).
- **Self-judgments excluded:** Models do not judge their own responses.
- **Weighted Score:** Composite of all 5 criteria.

---

## Citation

The Multivac (2026). Blind Peer Evaluation: CODE-010. app.themultivac.com

## License

Open data. Free to use, share, and build upon. Please cite The Multivac when using this data.

Download raw JSON: https://app.themultivac.com/api/evaluations/EVAL-20260207-142929/results
Full dataset: https://app.themultivac.com/dashboard/export