Stima delle capacità basata su più benchmark. Punti di indice, non una percentuale; gli intervalli mostrano l’incertezza.
130160
DeepSeek V3.2
DeepSeek-V3.2 · Intervallo 95% 144,4–147,5
Preferenze umane · Arena
Preferenze complessive sul testo da confronti diretti. Non misura accuratezza fattuale o riuscita dei compiti.
14101440
DeepSeek V3.2
DeepSeek-V3.2 · Intervallo 95% 1421,0–1427,9 · 48,095 voti · 2026-10-28
Risultati registrati
Ogni riga usa lo stesso benchmark, versione, carico, scenario e unità. I risultati mancanti non valgono zero. I risultati di sistema includono tutta la configurazione hardware e software.
Arena text preference / chinesehuman preference · 2026-10-28 · rating
DeepSeek V3.2
1.454,94 rating
Intervallo di confidenza: 1.444,89 – 1.464,99
3726 voti
Categoria
chinese
Configurazione valutata
DeepSeek-V3.2
Tipo di valutazione
human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating
DeepSeek V3.2
1.448,79 rating
Intervallo di confidenza: 1.442,56 – 1.455,03
10917 voti
Categoria
coding
Configurazione valutata
DeepSeek-V3.2
Tipo di valutazione
human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating
DeepSeek V3.2
1.399,59 rating
Intervallo di confidenza: 1.392,03 – 1.407,14
6873 voti
Categoria
creative_writing
Configurazione valutata
DeepSeek-V3.2
Tipo di valutazione
human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating
DeepSeek V3.2
1.436,64 rating
Intervallo di confidenza: 1.431,64 – 1.441,63
19604 voti
Categoria
english
Configurazione valutata
DeepSeek-V3.2
Tipo di valutazione
human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating
DeepSeek V3.2
1.414,57 rating
Intervallo di confidenza: 1.409,72 – 1.419,42
34020 voti
Categoria
exclude_ties
Configurazione valutata
DeepSeek-V3.2
Tipo di valutazione
human preference
Arena text preference / experthuman preference · 2026-10-28 · rating
DeepSeek V3.2
1.435,80 rating
Intervallo di confidenza: 1.425,60 – 1.446,01
3429 voti
Categoria
expert
Configurazione valutata
DeepSeek-V3.2
Tipo di valutazione
human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating
DeepSeek V3.2
1.429,05 rating
Intervallo di confidenza: 1.411,39 – 1.446,70
1241 voti
Categoria
french
Configurazione valutata
DeepSeek-V3.2
Tipo di valutazione
human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating
DeepSeek V3.2
1.408,12 rating
Intervallo di confidenza: 1.387,97 – 1.428,26
901 voti
Categoria
german
Configurazione valutata
DeepSeek-V3.2
Tipo di valutazione
human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating
DeepSeek V3.2
1.433,51 rating
Intervallo di confidenza: 1.429,08 – 1.437,94
27074 voti
Categoria
hard_prompts
Configurazione valutata
DeepSeek-V3.2
Tipo di valutazione
human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating
DeepSeek V3.2
1.445,45 rating
Intervallo di confidenza: 1.439,24 – 1.451,66
11105 voti
Categoria
hard_prompts_english
Configurazione valutata
DeepSeek-V3.2
Tipo di valutazione
human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.394,65 rating
Intervallo di confidenza: 1.387,84 – 1.401,45
8785 voti
Categoria
industry_entertainment_and_sports_and_media
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.424,97 rating
Intervallo di confidenza: 1.414,81 – 1.435,14
3491 voti
Categoria
industry_legal_and_government
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.445,47 rating
Intervallo di confidenza: 1.438,29 – 1.452,64
7632 voti
Categoria
industry_life_and_physical_and_social_science
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.438,92 rating
Intervallo di confidenza: 1.426,78 – 1.451,06
2483 voti
Categoria
industry_mathematical
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.439,77 rating
Intervallo di confidenza: 1.428,91 – 1.450,63
3118 voti
Categoria
industry_medicine_and_healthcare
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.443,67 rating
Intervallo di confidenza: 1.438,44 – 1.448,90
17256 voti
Categoria
industry_software_and_it_services
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.406,94 rating
Intervallo di confidenza: 1.400,66 – 1.413,22
10687 voti
Categoria
industry_writing_and_literature_and_language
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.412,63 rating
Intervallo di confidenza: 1.406,94 – 1.418,32
13567 voti
Categoria
instruction_following
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / japanesehuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.358,14 rating
Intervallo di confidenza: 1.329,82 – 1.386,46
452 voti
Categoria
japanese
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / koreanhuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.369,73 rating
Intervallo di confidenza: 1.348,50 – 1.390,97
796 voti
Categoria
korean
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / longer_queryhuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.428,02 rating
Intervallo di confidenza: 1.422,30 – 1.433,73
13732 voti
Categoria
longer_query
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / mathhuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.435,18 rating
Intervallo di confidenza: 1.424,60 – 1.445,76
3141 voti
Categoria
math
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.426,95 rating
Intervallo di confidenza: 1.419,90 – 1.434,00
8341 voti
Categoria
multi_turn
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.408,85 rating
Intervallo di confidenza: 1.404,54 – 1.413,17
28490 voti
Categoria
non_english
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.424,46 rating
Intervallo di confidenza: 1.420,99 – 1.427,92
48112 voti
Categoria
overall
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / polishhuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.408,46 rating
Intervallo di confidenza: 1.392,34 – 1.424,58
1340 voti
Categoria
polish
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / russianhuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.423,28 rating
Intervallo di confidenza: 1.415,57 – 1.431,00
6096 voti
Categoria
russian
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Arena text preference / spanishhuman preference · 2026-10-15 · rating
DeepSeek V3.2
1.422,91 rating
Intervallo di confidenza: 1.407,20 – 1.438,63
1531 voti
Categoria
spanish
Configurazione valutata
deepseek-v3.2
Tipo di valutazione
human preference
Esecuzione locale
Inferenza · stima della memoria
DeepSeek V3.2
Informazioni verificate sull’architettura insufficienti per una stima affidabile della VRAM.
Costo API
USD / mese
Stima per token di testo standard. Esclude imposte, cache, sconti batch e costi aggiuntivi per ragionamento e modalità multimediali. Limiti e disponibilità possono variare per provider.
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
input cost per token
6,2e-7*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
input cost per token
2,8e-7*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
input cost per token batches
3,1e-7*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
input cost per token cache hit
2,8e-8*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
input cost per token flex
3,1e-7*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
input cost per token priority
1,085e-6*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
litellm provider
deepseek*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
litellm provider
bedrock_converse*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
max input tokens
163.840*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
max input tokens
164.000*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
max output tokens
163.840*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
max output tokens
8.000*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
max tokens
8.000*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
max tokens
163.840*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
mode
chat*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
mode
chat*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
output cost per token
1,85e-6*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
output cost per token
4,0e-7*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
output cost per token batches
9,25e-7*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
output cost per token flex
9,25e-7*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
output cost per token priority
3,2375e-6*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports assistant prefill
Sì*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports audio input
No*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports function calling
Sì*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports function calling
Sì*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports native structured output
Sì*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports prompt caching
Sì*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports reasoning
Sì*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports reasoning
Sì*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports response schema
Sì*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports tool choice
Sì*Ambito e condizioni
native_context
provider
deepseek*
mode
chat*
full_native_model_id
deepseek/deepseek-v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports tool choice
Sì*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*
supports vision
No*Ambito e condizioni
native_context
provider
bedrock_converse*
mode
chat*
full_native_model_id
deepseek.v3.2*
pricing_scope
Literal source field names and values retained; provider, region, cache/batch/threshold conditions and per-unit meaning must be checked. max_tokens is not automatically a context limit.*