Benchmarks · agentic
DeepFinance
Julius Research's ultra long-horizon benchmark of real Private Equity and Investment Banking workflows: 21 tasks that take a professional about 27.1 hours each, scored against 84 rubric criteria per task, 10 attempts per task. Value is the mean share of rubric criteria met. Cost is mean API cost per task.
16 models · 16 results · updated 2026-11-05 · juliusresearch.ai ↗
Leaderboard
Opus 5.5 max
Heavy line: best score at each price. Thin lines join one model's effort levels. Cost uses each model's median subscription plan, or API price with no plan data. 5 more models under All models in settings.
More from this benchmark
3 published measuresScore vs any published measure3 measures: sub-scores, splits, counts
6 more models under All models in settings.
Breakdown
darker is better within each column · click a column to sort| Model | ||
|---|---|---|
| Opus 5.5 max | 210.0 | 9.1 |
| Sonnet 5.5 max | 210.0 | 9 |
| GPT-6 Astra xhigh | 210.0 | 8.1 |
| GPT-6.1 Sol xhigh | 210.0 | 7.8 |
| Fable 5.1 max | 209.0 | 5.3 |
| GPT-6.1 Terra xhigh | 210.0 | 6.8 |
| Gemini 4 Argon high | 208.0 | 7 |
| Muse Spark 1.3 max | 210.0 | 4.6 |
| Kimi K3.1 max | 210.0 | 4.4 |
| Grok 4.7 xhigh | 210.0 | 4.4 |
| Qwen 3.8 Max xhigh | 210.0 | 4.2 |
| GLM 5.3 max | 206.0 | 3.8 |
| DeepSeek V4.1 Flash max | 210.0 | 3.9 |
| GPT-6 Luna xhigh | 210.0 | 4.5 |
| Mistral Large 4 high | 210.0 | 4.4 |
| DeepSeek V4 Pro max | 202.0 | 4.1 |
All results
16 rows| # | Model | Effort | DeepFinance | Subscription $ / task | API $ / task |
|---|---|---|---|---|---|
| 1 | Opus 5.5 | max | 54.8% ±9.1 | $0.13 | $8.63 |
| 2 | Sonnet 5.5 | max | 54.6% ±8.9 | $0.10 | $6.94 |
| 3 | GPT-6 Astra | xhigh | 54.2% ±8.1 | $0.19 | $4.55 |
| 4 | GPT-6.1 Sol | xhigh | 52.9% ±7.8 | $0.047 | $0.81 |
| 5 | Fable 5.1 | max | 45.6% ±5.3 | $1.87 | $12 |
| 6 | GPT-6.1 Terra | xhigh | 43.5% ±6.8 | $0.019 | $0.38 |
| 7 | Gemini 4 Argon | high | 41.3% ±7.0 | $2.03 | $2.85 |
| 8 | Muse Spark 1.3 | max | 37.0% ±4.6 | $2.44 | $4.87 |
| 9 | Kimi K3.1 | max | 34.6% ±4.4 | $2.67 | $2.67 |
| 10 | Grok 4.7 | xhigh | 34.1% ±4.4 | $0.44 | $4.71 |
| 11 | Qwen 3.8 Max | xhigh | 30.3% ±4.2 | $0.59 | $4.99 |
| 12 | GLM 5.3 | max | 27.7% ±3.8 | $0.12 | $1.11 |
| 13 | DeepSeek V4.1 Flash | max | 27.3% ±3.8 | $0.009 | $0.13 |
| 14 | GPT-6 Luna | xhigh | 24.6% ±4.5 | $0.006 | $0.085 |
| 15 | Mistral Large 4 | high | 21.5% ±4.3 | $0.18 | $3.69 |
| 16 | DeepSeek V4 Pro | max | 20.5% ±4.1 | $0.25 | $1.50 |