# The Multivac — Evaluation Report

**Evaluation ID:** EVAL-20260403-102352
**Date:** Apr 03, 2026
**Category:** analysis
**Question ID:** ANALYSIS-021

---

## Question

You receive a job offer: $150K base, $50K RSUs/year, $20K signing bonus. The company's Glassdoor shows similar roles at $140K-$180K. You have a competing offer at $165K base, no RSUs. (1) Calculate total compensation for both over 4 years (assume 10% annual RSU appreciation). (2) What non-financial factors should affect the decision? (3) Write the counter-offer email that maximizes your outcome without risking the offer.

---

## Winner

**GPT-5.4** (openrouter)
- Winner Score: 9.11
- Matrix Average: 7.77
- Total Judgments: 81

---

## Rankings

| Rank | Model | Provider | Avg Score | Judgments |
|------|-------|----------|-----------|----------|
| 1 | GPT-5.4 | openrouter | 9.11 | 9 |
| 2 | Claude Sonnet 4.6 | openrouter | 8.79 | 9 |
| 3 | Grok 4.20 | openrouter | 8.68 | 9 |
| 4 | MiMo-V2-Flash | Xiaomi | 8.63 | 9 |
| 5 | Gemini 3 Flash Preview | Google | 8.30 | 9 |
| 6 | GPT-OSS-120B | OpenAI | 7.94 | 9 |
| 7 | DeepSeek V4 | openrouter | 7.36 | 9 |
| 8 | Claude Opus 4.6 | openrouter | 7.34 | 9 |
| 9 | Gemini 3.1 Pro | openrouter | 3.82 | 9 |

---

## 10×10 Judgment Matrix

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

| Judge ↓ / Resp → | DeepSeek V4 | Gemini 3.1 Pro | Claude Opus | GPT-5.4 | MiMo-V2-Flash | Claude Sonnet | Grok 4.20 | GPT-OSS-120B | Gemini 3 | MiniMax M2.5 |
|---|---|---|---|---|---|---|---|---|---|---|
| DeepSeek V4 | — | 6.2 | 9.2 | 9.2 | 8.8 | 8.8 | 8.8 | 8.8 | 8.8 | · |
| Gemini 3.1 Pro | 6.8 | — | 6.0 | 9.1 | 9.6 | 10.0 | 8.6 | 5.5 | 8.1 | · |
| Claude Opus | 7.0 | 2.6 | — | 8.9 | 8.5 | 8.5 | 8.3 | 8.0 | 8.1 | · |
| GPT-5.4 | 5.9 | 2.6 | 5.0 | — | 7.8 | 8.3 | 8.6 | 5.4 | 7.9 | · |
| MiMo-V2-Flash | 8.8 | 3.6 | 8.6 | 9.6 | — | 9.3 | 9.0 | 9.0 | 9.0 | · |
| Claude Sonnet | 7.3 | 3.0 | 8.1 | 9.1 | 8.6 | — | 8.6 | 8.6 | 8.3 | · |
| Grok 4.20 | 6.4 | 4.6 | 7.2 | 8.2 | 7.8 | 7.5 | — | 8.8 | 7.8 | · |
| GPT-OSS-120B | 8.0 | 2.6 | 5.0 | 8.8 | 8.6 | 8.6 | 7.5 | — | 8.0 | · |
| Gemini 3 | 8.8 | 6.0 | 9.4 | 9.8 | 9.8 | 9.8 | 9.6 | 9.0 | — | · |
| MiniMax M2.5 | 7.4 | 3.0 | 7.5 | 9.3 | 8.3 | 8.3 | 9.3 | 8.6 | 8.8 | — |

---

## 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: ANALYSIS-021. 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-20260403-102352/results
Full dataset: https://app.themultivac.com/dashboard/export