Is gemini-3.1-pro-preview-high good at Investor & Pitch?
gemini-3.1-pro-preview-high ranks #5 of 50 for Investor & Pitch — strong. The top pick for this task is claude-sonnet-4.6-high.
gemini-3.1-pro-preview-high on each Investor & Pitch sub-task
| Investor Question Test | 87.7/100 | #5 |
| Founder Reality Check | 87.0/100 | #6 |
| Market Sizing Reality Check | 85.7/100 | #12 |
| Deck Doctor | 81.7/100 | #17 |
Real examples, graded
WinEnterprise self-serve fantasy 92/100
“The response is expert-level, demonstrating a deep, practical understanding of enterprise SaaS realities, InfoSec blockers, and procurement dynamics. The proposed tests are highly actionable, and the tactful script perfectly balances validation with necessary reality checks. It is nearly production-ready as an advisory response.”
WinOverstuffed traction slide 89/100
“The model provides an expert-level response. It perfectly identifies the glaring red flag in the data (42 pilots vs 4 paying customers), distills the copy to the most impactful metrics, and correctly advises moving integrations to a product slide. The only minor flaw is including a meta-note within the copy block itself, though it still easily passes the word count constraint.”
Frequently asked
Is gemini-3.1-pro-preview-high good at Investor & Pitch?
gemini-3.1-pro-preview-high ranks #5 of 50 models we tested for Investor & Pitch, scoring strong.
What is gemini-3.1-pro-preview-high's strongest Investor & Pitch skill?
Its best sub-task here is Investor Question Test.
This page is Spring Prompt, running
We just did this for every model. Do it for your prompt.
The rankings above come from running real tasks through real models and scoring every output. Spring Prompt is that same engine — pointed at your prompt, your test cases, and your definition of good.
- Generate test cases from your prompt — no eval set required to start.
- Compare models side by side with quality, cost and latency in one matrix.
- Optimise the winner until the scores say it's ready to ship.
Prompt × model results
12 test cases · 3 evals