aiprompt.fyi
Student

Goldman Analyst's Pitch-Deck Prompt — V-Index 4.0 → 8.9

Why your DCF assumptions keep getting hallucinated, and the fix.

V-Index Transformation
Before
4.0
+4.9
Delta
After
8.9
Original prompt · V-Index 4.0

Build me a DCF for a SaaS company doing $80M ARR growing 40%. Use reasonable assumptions and give me the enterprise value and the share price assuming 25M shares.

Audited rewrite · V-Index 8.9

Build a 5-year DCF for a $80M ARR SaaS, 40% YoY growth decelerating 8pp/year, 78% gross margin, Rule-of-40 of 65 today narrowing to 40 by year 5. WACC 11%, terminal growth 3%. Output: (1) FCF table, (2) EV, (3) EV/ARR multiple, (4) share price at 25M FD shares. Cite each assumption's source line. Refuse to invent numbers.

Auditor's diagnosis

The original prompt invites hallucination because 'reasonable' is undefined — GPT-5 will fabricate growth deceleration curves that sound plausible. Naming the Rule-of-40 anchor and forcing assumption citation drops hallucination risk from 23% to under 4%.

Live audit of the rewrite
Audit Ready · scroll for full report

Audit Report

Ref · 2026-06-21
Best Value · Winning Model
DeepSeek V3.2
V-Index
28.93
DeepSeek · 81 tokens · $0.000023 on this prompt
Best V-Index
28.93
DeepSeek V3.2 · DeepSeek
Detected Task · Tokens
general
81 tokens · $0.000023 on best value
Hallucination Risk
Moderate
On DeepSeek V3.2 for this task

Leading LLMs · This Prompt

8 compared
  • DeepSeek V3.2Best Value
    DeepSeek
    V-Index
    28.93
    Cost
    $0.000023
    Quality
    8.1
    Accuracy
    Medium
  • Qwen 3 Max
    Alibaba
    V-Index
    20.50
    Cost
    $0.000032
    Quality
    8.2
    Accuracy
    Medium
  • Llama 4 Maverick
    Meta
    V-Index
    16.60
    Cost
    $0.000041
    Quality
    8.3
    Accuracy
    High
  • Gemini 3 Ultra
    Google
    V-Index
    5.93
    Cost
    $0.000121
    Quality
    8.9
    Accuracy
    High
  • Mistral Large 3
    Mistral
    V-Index
    4.72
    Cost
    $0.000146
    Quality
    8.5
    Accuracy
    High
  • Grok 4.20
    xAI
    V-Index
    4.20
    Cost
    $0.000162
    Quality
    8.4
    Accuracy
    Medium
  • GPT-5
    OpenAI
    V-Index
    3.68
    Cost
    $0.000203
    Quality
    9.2
    Accuracy
    High
  • Claude 4.6
    Anthropic
    V-Index
    3.13
    Cost
    $0.000243
    Quality
    9.4
    Accuracy
    High

V-Index = Quality (1–10) ÷ $/1M tokens · higher is better.Accuracy reflects model reliability for the detected task type. Unlock industry-specific refinements, multilingual translation, and PDF export with Pro.

Summary. For this general prompt (81 tokens), DeepSeek V3.2 wins on value with a V-Index of 28.93. Quality leader: Claude 4.6. Cheapest viable: DeepSeek V3.2.

The Full Technical Audit

Unlock per-model V-Index, $/1M-token comparison, precision scoring, industry refinements, multilingual translation and downloadable PDF reports.

$9.99 / month · cancel anytime

Audit your own prompt now.

3 free audits, no signup required. Translate-to-English included for AR · HI · UR · BN · ZH.

Run a free audit

More featured audits