Official mode label: Non-Think; the row-specific inference budget and sampling parameters are not stated in this model-card table.
未知项
Do not assume local recommended parameters are the evaluation settings.
工具
工具条件未记录
推理预算
未记录
框架
未记录
协议
Official DeepSeek-V4 report Table 7; row-specific internal harness, tools and context-management protocol retained. The table does not print percent signs unless the metric itself is Elo.
条件
nonthink mode; Chinese-SimpleQA row. This value is explicitly in the named V4 preview mode column and is not propagated to later dated revisions.
Official mode label: High; the row-specific inference budget and sampling parameters are not stated in this model-card table.
未知项
Do not assume local recommended parameters are the evaluation settings.
工具
工具条件未记录
推理预算
High
框架
未记录
协议
Official DeepSeek-V4 report Table 7; row-specific internal harness, tools and context-management protocol retained. The table does not print percent signs unless the metric itself is Elo.
条件
high mode; Chinese-SimpleQA row. This value is explicitly in the named V4 preview mode column and is not propagated to later dated revisions.
Official mode label: Max; the row-specific inference budget and sampling parameters are not stated in this model-card table.
未知项
Do not assume local recommended parameters are the evaluation settings.
工具
工具条件未记录
推理预算
Max
框架
未记录
协议
Official DeepSeek-V4 report Table 7; row-specific internal harness, tools and context-management protocol retained. The table does not print percent signs unless the metric itself is Elo.
条件
max mode; Chinese-SimpleQA row. This value is explicitly in the named V4 preview mode column and is not propagated to later dated revisions.
DeepSeek-V4-Pro (Preview) · Chinese-SimpleQA · DeepSeek-V4 report Table 7; Chinese-SimpleQA (Pass@1); unit not printed · Scores and configurations
DeepSeek-V4-Pro (Preview) record for Chinese-SimpleQA · DeepSeek-V4 report Table 7; Chinese-SimpleQA (Pass@1); unit not printed. 3 score records for this release, including 3 exact values (may include historical records). Includes 3 source locations.
DeepSeek-V4-Pro (Preview) · Chinese-SimpleQA scores and configurations
Metric and score
Configuration and evaluation conditions
Sources
Pass@175.8Reported 75.8 · Unit not specified
Published Unconfirmed
DeepSeek-V4-Pro (Preview) Non-Think
Settings
Official mode label: Non-Think; the row-specific inference budget and sampling parameters are not stated in this model-card table.
Unknowns
Do not assume local recommended parameters are the evaluation settings.
Tools
Tool access not recorded
Reasoning budget
not recorded
Framework
not recorded
Protocol
Official DeepSeek-V4 report Table 7; row-specific internal harness, tools and context-management protocol retained. The table does not print percent signs unless the metric itself is Elo.
Conditions
nonthink mode; Chinese-SimpleQA row. This value is explicitly in the named V4 preview mode column and is not propagated to later dated revisions.
Official mode label: High; the row-specific inference budget and sampling parameters are not stated in this model-card table.
Unknowns
Do not assume local recommended parameters are the evaluation settings.
Tools
Tool access not recorded
Reasoning budget
High
Framework
not recorded
Protocol
Official DeepSeek-V4 report Table 7; row-specific internal harness, tools and context-management protocol retained. The table does not print percent signs unless the metric itself is Elo.
Conditions
high mode; Chinese-SimpleQA row. This value is explicitly in the named V4 preview mode column and is not propagated to later dated revisions.
Official mode label: Max; the row-specific inference budget and sampling parameters are not stated in this model-card table.
Unknowns
Do not assume local recommended parameters are the evaluation settings.
Tools
Tool access not recorded
Reasoning budget
Max
Framework
not recorded
Protocol
Official DeepSeek-V4 report Table 7; row-specific internal harness, tools and context-management protocol retained. The table does not print percent signs unless the metric itself is Elo.
Conditions
max mode; Chinese-SimpleQA row. This value is explicitly in the named V4 preview mode column and is not propagated to later dated revisions.