Finance Agent v2 — Results
| Rank | Model | Score |
|---|---|---|
| 1 | kimi-k3 | 54.4 |
Finance Agent v2
1 phaseActive
Second-generation financial agent benchmark testing multi-step finance task automation. Metric: % task success.
Overview
Finance Agent v2
Quick answer: Finance Agent v2 is a second-generation benchmark testing AI agents on multi-step financial task automation — workflows that combine data retrieval, calculation, and financial decision-making rather than single-turn Q&A. Kimi K3 scores 54.4% as of July 2026.
At a Glance
What it tests: An agent's ability to complete multi-step financial workflows end-to-end, combining data retrieval, computation, and financial reasoning.
Why it matters: Financial task automation (e.g., reconciliation, reporting, analysis pipelines) is a high-value enterprise use case where agentic reliability matters more than single-turn accuracy.
Known limitations: As a "v2" iteration, exact differences from any prior version and detailed task composition are not yet independently published outside its citation by Moonshot AI.
What Finance Agent v2 Measures
Finance Agent v2 evaluates an agent's ability to complete multi-step financial workflows — such as retrieving financial data, performing calculations, and producing analysis outputs — testing agentic reliability across the full task pipeline rather than isolated question-answering.
Benchmark Specifications
| Field | Value |
|---|---|
| Task category | Agent / finance automation |
| Metric | % task success |
| Saturation | Low |
| Created by | Not yet independently documented |
How Finance Agent v2 Is Scored
Agents complete multi-step financial workflows, with outcomes verified against expected results, producing an aggregate % task success rate.
State-of-the-Art Results
| Rank | Model | Score | Source | Date |
|---|---|---|---|---|
| 1 | Kimi K3 | 54.4% | Kimi K3 technical report | 2026-07 |
Score sourced from Moonshot AI's Kimi K3 announcement, July 2026.
Finance Agent v2 on Benchgen
No Benchgen results yet — be the first to run Finance Agent v2.
Finance Agent v2 vs Other Benchmarks
| Benchmark | What it tests | Saturation |
|---|---|---|
| Finance Agent v2 | Financial task automation workflows | Low |
| CorpFin v2 | Corporate finance analysis & modeling | Low |
| SpreadsheetBench 2 | Spreadsheet manipulation & formula tasks | Low |
| SaaS-Bench | SaaS product usage workflows | Low |
Run Finance Agent v2 on Your Model
Benchgen lets you run Finance Agent v2 against your own model, tracking financial workflow automation task success over time.