Skip to content

Benchmark

FinBen ccf F1

Scores are percentages. Higher is better.

Models measured
15
Scores recorded
15
Sources
1
IQ weight
0.47

Score distribution

44.8%
Mean
24.1%
Median
40.6%
Std dev
0%
10th percentile
99.8%
90th percentile
0%
Min
0%99.8%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.47 for chat models, measured over 15 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1codellama-7bpublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results99.8%
2Llama 3.2 3Bpublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results99.8%
3Qwen2.5 7Bpublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results99.8%
4Qwen2.5 7B Instructpublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results97.3%
5gemini-1.0-propublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results96%
6GPT-4 Turbopublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results55%
7DeepSeek V3published-demo-leaderboard, ccf, F1· TheFinAI public numeric results41.1%
8DeepSeek R1 Distill Qwen 7Bpublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results24.1%
9GPT-3.5 Turbopublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results20%
10Llama 2 70bpublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results17%
11Qwen2 7B Instructpublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results13%
12Mistral (7B) Instruct v0.3published-demo-leaderboard, ccf, F1· TheFinAI public numeric results9.3%
13code-llama-7b-instructpublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results0%
14Llama-2-7b-chat-hfpublished-demo-leaderboard, ccf, F1· TheFinAI public numeric results0%
15Mixtral 8X7b V0.1published-demo-leaderboard, ccf, F1· TheFinAI public numeric results0%

Data credits: The Fin AI. Published FinBen result summaries and Open Greek Financial LLM Leaderboard aggregates. Results licence not separately stated; fractional FinBen success metrics converted to percentages.