AI BENCHY Compare
DeepSeek: DeepSeek V3.2 vs Inception: Mercury 2
Linganisha:
Benchmark zimetengenezwa kutoka seti za majaribio za AI BENCHY tarehe: 2026-03-05
| Kipimo | DeepSeek: DeepSeek V3.2 none Toleo: 2025-12-01 | Inception: Mercury 2 medium Toleo: 2026-02-24 |
|---|---|---|
| Nafasi | #38 | #35 |
| Wastani wa alama | 52 | 54 |
| Uthabiti | 83 | 83 |
| Gharama kwa matokeo | 0.254 | 0.622 |
| Jumla ya gharama | $0.016 | $0.044 |
| Majaribio sahihi | ||
| Kiwango cha kupita kwa kila jaribio | 51.1% | 57.8% |
| Majaribio yasiyo thabiti | 3 | 3 |
| Tokeni za matokeo | 7,756 | 3,571 |
| Tokeni za hoja | 0 | 45,379 |
Alama dhidi ya gharama ya jumla
Mgawanyo wa kategoria
| Mbinu za kupinga AI | Alama | Uthabiti | Kiwango cha kupita kwa kila jaribio | Majaribio yasiyo thabiti | Majaribio sahihi | Tokeni za matokeo | Tokeni za hoja |
|---|---|---|---|---|---|---|---|
| DeepSeek: DeepSeek V3.2 | 100 | 97 | 0.0% | 0 | 1,411 | 0 | |
| Inception: Mercury 2 | 73 | 98 | 66.7% | 0 | 2,531 | 2,410 |
| Mchanganyiko | Alama | Uthabiti | Kiwango cha kupita kwa kila jaribio | Majaribio yasiyo thabiti | Majaribio sahihi | Tokeni za matokeo | Tokeni za hoja |
|---|---|---|---|---|---|---|---|
| DeepSeek: DeepSeek V3.2 | 80 | 100 | 0.0% | 0 | 2,887 | 0 | |
| Inception: Mercury 2 | 100 | 100 | 100.0% | 0 | 268 | 4,887 |
| Uchanganuzi na uchimbaji wa data | Alama | Uthabiti | Kiwango cha kupita kwa kila jaribio | Majaribio yasiyo thabiti | Majaribio sahihi | Tokeni za matokeo | Tokeni za hoja |
|---|---|---|---|---|---|---|---|
| DeepSeek: DeepSeek V3.2 | 54 | 58 | 66.7% | 1 | 1,710 | 0 | |
| Inception: Mercury 2 | 55 | 59 | 83.3% | 1 | 183 | 1,656 |
| Mahususi kwa domeni | Alama | Uthabiti | Kiwango cha kupita kwa kila jaribio | Majaribio yasiyo thabiti | Majaribio sahihi | Tokeni za matokeo | Tokeni za hoja |
|---|---|---|---|---|---|---|---|
| DeepSeek: DeepSeek V3.2 | 100 | 72 | 22.2% | 1 | 24 | 0 | |
| Inception: Mercury 2 | 100 | 72 | 11.1% | 1 | 41 | 30,754 |
| Ufuataji wa maagizo | Alama | Uthabiti | Kiwango cha kupita kwa kila jaribio | Majaribio yasiyo thabiti | Majaribio sahihi | Tokeni za matokeo | Tokeni za hoja |
|---|---|---|---|---|---|---|---|
| DeepSeek: DeepSeek V3.2 | 100 | 100 | 100.0% | 0 | 66 | 0 | |
| Inception: Mercury 2 | 100 | 100 | 100.0% | 0 | 14 | 958 |
| Puzzle Solving | Alama | Uthabiti | Kiwango cha kupita kwa kila jaribio | Majaribio yasiyo thabiti | Majaribio sahihi | Tokeni za matokeo | Tokeni za hoja |
|---|---|---|---|---|---|---|---|
| DeepSeek: DeepSeek V3.2 | 77 | 75 | 88.9% | 1 | 1,136 | 0 | |
| Inception: Mercury 2 | 17 | 75 | 22.2% | 1 | 354 | 2,758 |
| Mwito wa zana | Alama | Uthabiti | Kiwango cha kupita kwa kila jaribio | Majaribio yasiyo thabiti | Majaribio sahihi | Tokeni za matokeo | Tokeni za hoja |
|---|---|---|---|---|---|---|---|
| DeepSeek: DeepSeek V3.2 | 100 | 100 | 100.0% | 0 | 522 | 0 | |
| Inception: Mercury 2 | 100 | 100 | 100.0% | 0 | 180 | 1,956 |
Ulinganisho wa haraka
Badilisha jozi ya ulinganisho
DeepSeek V3.2nonevsgpt-oss-120bmediumInapatikana bureGemini 2.5 FlashnonevsMercury 2mediumMercury 2mediumvsQwen3.5-122B-A10BnoneMercury 2mediumvsGLM 5noneDeepSeek V3.2nonevsMiniMax M2.5mediumDeepSeek V3.2nonevsGPT-5 NanomediumMercury 2mediumvsQwen3.5-FlashnoneMercury 2mediumvsQwen3.5-27BnoneDeepSeek V3.2nonevsQwen3.5-35B-A3BmediumMercury 2mediumvsQwen3.5-35B-A3BnoneMercury 2mediumvsGPT-5.4noneDeepSeek V3.2nonevsGPT-5 Minimedium