Fiction.liveBench ↗52.8%
DeepSeek-V3.1 · Fiction.LiveBench · 16k token score · % ↗
DeepSeek / MODEL EVIDENCE
DeepSeek V3.1
12 benchmarks · 13 observations · 12 sources · released 2025-08-21
Scores at a glance
12 benchmarks · large dot: DeepSeek V3.1; small dots: closest models where reported
DeepSeek-V3.1 · SimpleBench · Score (AVG@5) · % ↗
WeirdML (v2) ↗38.36%
DeepSeek-V3.1 · WeirdML · Accuracy · % ↗
1 more reported result
37.5% · DeepSeek-V3.1 · WeirdML · Accuracy · % · Epoch AI · WeirdML ↗
LMCA ↗24.29%
DeepSeek-V3.1 · LMCA · Score · % ↗
DTBench ↗82.67%
DeepSeek-V3.1 · DTBench · Accuracy · % ↗
Epoch Capabilities Index (ECI) ↗139.79 ECI
DeepSeek-V3.1 · General ECI · 26 October 2026 fit ↗
CaseLaw v2 ↗53.91 %
DeepSeek V3.1 · CaseLaw · Accuracy ↗
CorpFin v2 ↗51.48 %
DeepSeek V3.1 · CorpFin · Accuracy ↗
Poker Agent Benchmark ↗1058.23 TrueSkill
DeepSeek V3.1 · Poker Agent · TrueSkill Rating ↗
Fortress ↗60.55%
DeepSeek-V3.1 · Scale Labs · Risk elicitation score ↗
DeepSeek-V3.1 · Scale Labs · Score ↗
ForecastBench ↗60.7 Brier index
deepseek-v3.1† · Overall difficulty-adjusted Brier index ↗
Closest models by score
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
DeepSeek V3.2 →DeepSeek
3 of 5 shared benchmark comparisons
Gemini 3 Pro (11/25) →Google
3 of 5 shared benchmark comparisons
GPT OSS 120B →OpenAI
3 of 5 shared benchmark comparisons
How similarity is measured
Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.
Sources
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
- DTBench — Conceptual Reasoning Index ↗
- LMCA — Conceptual Reasoning Index ↗
- Forecastbench — methodology and published results ↗
- SimpleBench ↗
- AI benchmarks & capabilities ↗
- Fiction.liveBench — March 14, 2025 benchmark release ↗
- WeirdML ↗
- Scale Labs — evaluation methodology and results ↗
- Scale Labs — evaluation methodology and results ↗
- CaseLaw — evaluation methodology and results ↗
- CorpFin — evaluation methodology and results ↗
- Poker Agent — evaluation methodology and results ↗
Names used in source tables
DeepSeek V3.1 · DeepSeek-V3.1 · deepseek-v3.1†