GPT-5 Mini
VS
MiniMax M2.7
GPT-5 Mini vs MiniMax M2.7: which wins at real work?
22 task areas · same graded test runs · rank comparison only, so 0–100 and Elo collections never mix raw scores.
GPT-5 Mini wins 21 of 22 task areas we tested; MiniMax M2.7 takes 1. MiniMax M2.7 costs 2.5× less per token ($0.9 vs $2.25 per 1M).
MiniMax M2.7 costs 2.5× less per token ($0.9 vs $2.25 per 1M).
Task by task
| Task area | GPT-5 Mini | MiniMax M2.7 | Winner |
|---|---|---|---|
| Chef / Home Cooking |
#45
/ 126
Usable
|
#124
/ 126
Weak
|
GPT-5 Mini |
| Frontend & Landing Pages |
#12
/ 109
Needs editing
|
#89
/ 109
Weak
|
GPT-5 Mini |
| Research & Competitive Analysis |
#21
/ 110
Excellent
|
#96
/ 110
Needs editing
|
GPT-5 Mini |
| Legal & HR |
#10
/ 110
Excellent
|
#81
/ 110
Strong
|
GPT-5 Mini |
| Coding |
#37
/ 115
Strong
|
#105
/ 115
Usable
|
GPT-5 Mini |
| Knowledge & Docs |
#41
/ 110
Strong
|
#108
/ 110
Weak
|
GPT-5 Mini |
| Data & Analytics |
#27
/ 110
Excellent
|
#92
/ 110
Excellent
|
GPT-5 Mini |
| Product & Project Management |
#1
/ 110
Excellent
|
#64
/ 110
Strong
|
GPT-5 Mini |
| Presentations & Decks |
#39
/ 110
Excellent
|
#83
/ 110
Strong
|
GPT-5 Mini |
| Creative & Comedy | #70 / 110 | #108 / 110 | GPT-5 Mini |
| Translation & Localization |
#52
/ 110
Excellent
|
#82
/ 110
Excellent
|
GPT-5 Mini |
| Content & Brand |
#69
/ 124
Strong
|
#96
/ 124
Usable
|
GPT-5 Mini |
| Training & Education |
#81
/ 110
Strong
|
#106
/ 110
Usable
|
GPT-5 Mini |
| Summarization & Meeting Notes |
#81
/ 110
Excellent
|
#104
/ 110
Strong
|
GPT-5 Mini |
| AI Strategy |
#100
/ 126
Usable
|
#119
/ 126
Usable
|
GPT-5 Mini |
| Landing Pages |
#50
/ 72
Usable
|
#68
/ 72
Needs editing
|
GPT-5 Mini |
| Investor & Pitch |
#50
/ 66
Usable
|
#66
/ 66
Weak
|
GPT-5 Mini |
| RAG, Safety & Grounding |
#78
/ 113
Excellent
|
#94
/ 113
Strong
|
GPT-5 Mini |
| Structured Output |
#17
/ 113
Excellent
|
#33
/ 113
Excellent
|
GPT-5 Mini |
| Sales |
#102
/ 110
Needs editing
|
#108
/ 110
Weak
|
GPT-5 Mini |
| Customer Support |
#104
/ 113
Usable
|
#108
/ 113
Needs editing
|
GPT-5 Mini |
| Executive Assistant |
#108
/ 112
Usable
|
#107
/ 112
Usable
|
MiniMax M2.7 |
Rank = position among every model config we tested in that task area (lower is better). Sorted by biggest gap first.
Same task, both models — judged
Both models answered the same test case; an independent judge graded each. Quotes are the judge's actual rationale.
Product & Project Management
Churn-risk surfacing (Northwind) (PRD / Spec)
“The PRD perfectly follows the instructions, leading with a clear problem statement and target user before any solutioning. It defines highly specific, measurable outcome metrics (including precision/recall targets for the model). It includes explicit non-goals, edge cases, and open questions without fabricating any user research.”
“No valid scores could be extracted from the malformed response.”
Frequently asked
Is GPT-5 Mini better than MiniMax M2.7?
Across 22 task areas we benchmarked, GPT-5 Mini ranks higher in 21 and MiniMax M2.7 in 1.
Which is cheaper, GPT-5 Mini or MiniMax M2.7?
MiniMax M2.7 costs 2.5× less per token ($0.9 vs $2.25 per 1M).
What is GPT-5 Mini better at?
GPT-5 Mini out-ranks MiniMax M2.7 at Chef / Home Cooking, Frontend & Landing Pages, Research & Competitive Analysis.
What is MiniMax M2.7 better at?
MiniMax M2.7 out-ranks GPT-5 Mini at Executive Assistant.
More comparisons
This page is Spring Prompt, running in public
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