GPT-6.1 Sol vs Mercury 3
Too close to call. On the work we tested, GPT-6.1 Sol and Mercury 3 can't be told apart. Mercury 3 is the cheaper one, by 3.6×.
GrindBench-PubSec
CGPT-6.1 SolOpenAI58% 52–63%1.3¢ a task · 300 tasksBMercury 3Inception61% 56–67%0.53¢ a task · 300 tasks
Too close to call here: their ranges overlap, so on another set of tasks the order could flip.
| Kind of work | GPT-6.1 Sol | Mercury 3 |
|---|---|---|
| ATIP redaction | 0% | 11% |
| Procurement evaluation | 80% | 86% |
| Performance analysis | 60% | 54% |
| Bilingual equivalence | 85% | 97% |
GrindBench-Banking
BGPT-6.1 SolOpenAI63% 50–75%5.4¢ a task · 60 tasksCMercury 3Inception52% 38–63%1.4¢ a task · 60 tasks
Too close to call here: their ranges overlap, so on another set of tasks the order could flip.
| Kind of work | GPT-6.1 Sol | Mercury 3 |
|---|---|---|
| Lending & mortgages | 60% | 50% |
| AML & KYC | 30% | 40% |
| Deposits & payments | 60% | 40% |
| Complaints & conduct | 100% | 70% |
| Credit risk | 60% | 60% |
| Capital & liquidity | 70% | 50% |
Each kind of work here rests on 10 tasks per model, so a single row can easily be 20 points off. A lead of 10 points or more is in bold; read it as a hint, not a result.