Muse Spark 1.3 vs DeepSeek V4.1 Flash
Too close to call. On the work we tested, Muse Spark 1.3 and DeepSeek V4.1 Flash can't be told apart. DeepSeek V4.1 Flash is the cheaper one, by 4.2×.
GrindBench-PubSec
BMuse Spark 1.3Meta64% 59–70%1.7¢ a task · 300 tasksBDeepSeek V4.1 FlashDeepSeek64% 59–70%0.43¢ 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 | Muse Spark 1.3 | DeepSeek V4.1 Flash |
|---|---|---|
| ATIP redaction | 25% | 20% |
| Procurement evaluation | 78% | 78% |
| Performance analysis | 61% | 63% |
| Bilingual equivalence | 93% | 95% |
GrindBench-Banking
BMuse Spark 1.3Meta63% 52–75%6.6¢ a task · 60 tasksBDeepSeek V4.1 FlashDeepSeek62% 48–73%1.6¢ 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 | Muse Spark 1.3 | DeepSeek V4.1 Flash |
|---|---|---|
| Lending & mortgages | 60% | 60% |
| AML & KYC | 70% | 70% |
| Deposits & payments | 40% | 40% |
| Complaints & conduct | 80% | 70% |
| Credit risk | 70% | 80% |
| Capital & liquidity | 60% | 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.