#12
DeepSeek V3.2
Deepseek ยท Toleo: 2025-12-01 ยท deepseek/deepseek-v3.2::medium
Wastani wa alama
6.99
Gharama kwa matokeo
0.193
Uthabiti
8.75
Jumla ya gharama
$0.018
Majaribio sahihi
9
Jaribio huhesabiwa kuwa limepita kikamilifu tu ikiwa run zake zote zimepita.
Majaribio yenye makosa
5
Kiwango cha kupita kwa kila jaribio: 71.4%
Majaribio yasiyo thabiti
2
Muda wa majibu: wastani 36.24s ยท jumla 507.30s ยท upeo 79.03s
Jibu lisilo sahihi: 3
Hakufuata maelekezo: 1
Muda umeisha: 1
Modeli bora kwa alama
Chagua modeli ya kwanza, kisha bofya modeli ya pili kufungua ukurasa wa kulinganisha bega kwa bega.
Ulinganisho wa haraka
DeepSeek V3.2mediumvsStep 3.5 FlashmediumInapatikana bureDeepSeek V3.2mediumvsGPT-5.2mediumDeepSeek V3.2mediumvsGemini 3 Flash PreviewnoneDeepSeek V3.2mediumvsQwen3.5-122B-A10BmediumDeepSeek V3.2mediumvsGLM 5mediumDeepSeek V3.2mediumvsClaude Sonnet 4.6noneDeepSeek V3.2mediumvsGemini 3 Flash PreviewmediumDeepSeek V3.2mediumvsGemini 3.1 Pro Previewmedium
Mgawanyo wa kategoria
| Kategoria | Wastani wa alama | Uthabiti | Majaribio sahihi |
|---|---|---|---|
| Anti-AI Tricks | 7.00 | 9.86 | 2/3 |
| Data parsing and extraction | 9.88 | 10.00 | 2/2 |
| Domain specific | 4.00 | 7.21 | 1/3 |
| Instructions following | 7.00 | 9.84 | 1/2 |
| Puzzle Solving | 7.00 | 7.21 | 2/3 |
| Tool Calling | 10.00 | 10.00 | 1/1 |