HOW TO READ THIS
Score is only one signal.
Consensus score estimates relative capability across public evaluations. Coverage tells you how much evidence sits behind the placement, while confidence describes evidence quality rather than intelligence.
← Back to Model Index
Summary snapshot
DeepSeek
DeepSeek V4 Pro 0813
Model evaluation profile · Source snapshot checked Aug 21, 2026
AI CENTER CONSENSUS
58.7Current rank #15Evaluation coverage73.0%Evidence currently available
Evidence confidenceMediumCompleteness, not model quality
Input price¥9.00Per 1M tokens
Output price¥27.00Per 1M tokens
SOURCE SCORECARD
How this model performs across evaluations.
Original benchmark rank, score, and model identifiers stay visible for auditability. Missing evaluations remain visible instead of being treated as zeroes.
Model evaluation summary
The aggregate evaluation is available below. Original benchmark providers are linked here so each result can be checked at its source.
Consensus58.7Coverage73.0%EvidenceMedium
MODEL SNAPSHOT
Release date2026-08-13
Current rank#15 / 30
Pricing basisCNY / 1M tokens
Benchmark results link directly to the original evaluation providers listed above.