# Best AI models for finance Rankings by use case and by benchmark. Every ranking names the benchmark it comes from and when the number was reported. Nothing is rolled into one score. ## Use cases - Best AI for Analyst Financial research (from Deep FinResearch Bench, Finance Agent Benchmark, FinDeepResearch, FinResearchBench, FrontierFinance, Vals Finance Index): https://benchmarks.finance/best/analyst-research/ - Best AI for Financial filings Q&A (from CorpFin v2, FinanceBench, FinanceQA, IPO Finance Agent, Moneda FilingsBench): https://benchmarks.finance/best/filings-qa/ - Best AI for Live financial market data (from CAIA (Crypto AI Agent Benchmark), FinSearchComp): https://benchmarks.finance/best/live-market-data/ - Best AI for Finance knowledge (from Advanced Financial Reasoning at Scale (CFA Level III evaluation), BizFinBench, Open FinLLM Leaderboard, Reasoning Models Ace the CFA Exams (CFA mock-exam evaluation), Tax Agent Bench): https://benchmarks.finance/best/finance-knowledge/ - Best AI for Financial Verifiability (from ClaimBench, FAITH, FinLFQA, K-FinHallu): https://benchmarks.finance/best/verifiability/ - Best AI for Financial advice and safety (from FinSafetyBench, FinTrust): https://benchmarks.finance/best/advice-and-safety/ - Best AI for Financial trading and portfolios (from Agent Market Arena (AMA), AI-Trader, InvestorBench, KTD-Fin, LiveTradeBench, Moneda PortfolioBench, PolyBench, PortBench, StockBench): https://benchmarks.finance/best/trading-and-portfolios/ - Best AI for Financial spreadsheets and accounting (from APEX-Accounting, BlueFin, Excel Modeling Benchmark (EMB), Finch: Finance & Accounting Cowork Benchmark): https://benchmarks.finance/best/spreadsheets-and-accounting/ ## By benchmark - AA-AnalystAgent (Financial spreadsheets and accounting): Gemini 3.7 Flash (High): 60 pass^5 % (Artificial Analysis evaluation page, date not stated), https://benchmarks.finance/leaderboards/aa-analyst-agent/ - Advanced Financial Reasoning at Scale (CFA Level III evaluation) (Finance knowledge): o4-mini: 79.1 overall % (mean of MCQ and essay) (Paper Table 1, 22 Sept 2025), https://benchmarks.finance/leaderboards/cfa-level-3-at-scale/ - Agent Market Arena (AMA) (Financial trading and portfolios): InvestorAgent / GPT-4.1: 40.8 cumulative return % (TSLA) (Paper Table 1, 30 Oct 2025), https://benchmarks.finance/leaderboards/agent-market-arena/ - AI-Trader (Financial trading and portfolios): MiniMax-M2: 9.6 cumulative return % (Paper Table 1, 1 Dec 2025), https://benchmarks.finance/leaderboards/ai-trader/ - APEX-Accounting (Financial spreadsheets and accounting): Gemini 4 Argon: 71.5 Mean Score (% criteria passed) (Mercor APEX-Accounting leaderboard, 1 Nov 2026), https://benchmarks.finance/leaderboards/apex-accounting/ - BizFinBench (Finance knowledge): ChatGPT-o3: 73.9 average score across 9 tasks (Paper Table 3, 26 May 2025), https://benchmarks.finance/leaderboards/bizfinbench/ - BlueFin (Financial spreadsheets and accounting): GPT 5.5: 49.6 criteria-weighted score % (overall) (Paper Table 2, May 2026), https://benchmarks.finance/leaderboards/bluefin/ - CAIA (Crypto AI Agent Benchmark) (Live financial market data): gpt-5: 67.4 accuracy % (majority vote, 5 runs) (Paper Tables 2-3, 17 Jan 2026), https://benchmarks.finance/leaderboards/caia/ - ClaimBench (Financial Verifiability): d1-ultra: 100 accuracy % (Vals.ai leaderboard, 23 Oct 2026), https://benchmarks.finance/leaderboards/vals-claimbench/ - CorpFin v2 (Financial filings Q&A): Claude Opus 5: 73.2 accuracy % (Vals.ai leaderboard (archived), 12 Aug 2026), https://benchmarks.finance/leaderboards/vals-corpfin-v2/ - Deep FinResearch Bench (Analyst Financial research): OpenAI: 86 overall factual claim rate % (Paper Table 7, 22 Apr 2026), https://benchmarks.finance/leaderboards/deep-finresearch-bench/ - Excel Modeling Benchmark (EMB) (Financial spreadsheets and accounting): Claude Fable 5.1: 76.7 accuracy % (Vals.ai leaderboard, 29 Oct 2026), https://benchmarks.finance/leaderboards/vals-emb/ - FAITH (Financial Verifiability): Claude Sonnet 4: 95.6 overall accuracy % (Paper Table 3, 24 Oct 2025), https://benchmarks.finance/leaderboards/faith/ - Finance Agent Benchmark (Analyst Financial research): Numa 1 (public set): 80.2 Partial Credit accuracy % (Moneda run, 2026-10-09, 9 Oct 2026), https://benchmarks.finance/leaderboards/vals-finance-agent/ - FinanceBench (Financial filings Q&A): Numa 1: 67 % correct, no filing given (150-question sample) (Moneda run, 2026-11-02, 2 Nov 2026), https://benchmarks.finance/leaderboards/financebench/ - FinanceQA (Financial filings Q&A): o1: 0.487 accuracy (total, 0-1) (Paper Table 1, 30 Jan 2025), https://benchmarks.finance/leaderboards/afterquery-financeqa/ - Finch: Finance & Accounting Cowork Benchmark (Financial spreadsheets and accounting): GPT 5.1 Pro: 38.4 workflow pass rate % (human eval) (Paper Table 5, Apr 2026), https://benchmarks.finance/leaderboards/finch/ - FinDeepResearch (Analyst Financial research): FinAgentTest: 41.9 overall accuracy score (0-100) (OpenFinArena leaderboard, 23 Mar 2026), https://benchmarks.finance/leaderboards/findeepresearch/ - FinLFQA (Financial Verifiability): GPT-4o: 56.5 evidence attribution F1 (Paper Table 6, 7 Oct 2025), https://benchmarks.finance/leaderboards/finlfqa/ - FinResearchBench (Analyst Financial research): Gemini: 76.6 Final Score (0-100) (Paper Table 1, 20 Oct 2025), https://benchmarks.finance/leaderboards/finresearchbench/ - FinSafetyBench (Financial advice and safety): LLaMA-3: 32.3 attack success rate % (financial crimes, English) (Paper Table 2, 1 May 2026), https://benchmarks.finance/leaderboards/finsafetybench/ - FinSearchComp (Live financial market data): grok-4: 68.9 Global subset average accuracy % (Paper Table 5, 16 Sept 2025), https://benchmarks.finance/leaderboards/finsearchcomp/ - FinTrust (Financial advice and safety): o4-mini: 1 safety harmfulness score (1-10, GPT-4.1 mini judge) (Paper Table 7 (App. A.2), 17 Oct 2025), https://benchmarks.finance/leaderboards/fintrust/ - FrontierFinance (Analyst Financial research): Samaya Sterling: 58.9 rubrics qualification rate % (Samaya leaderboard, 28 Oct 2026), https://benchmarks.finance/leaderboards/samaya-frontierfinance/ - InvestorBench (Financial trading and portfolios): Qwen2.5-72B-Instruct: 46.2 cumulative return % (Paper Table 2, 24 Dec 2024), https://benchmarks.finance/leaderboards/investorbench/ - IPO Finance Agent (Financial filings Q&A): GLM-5.2: 79.8 mean rubric score % (Paper Table 1, 30 Jun 2026), https://benchmarks.finance/leaderboards/ipo-finance-agent/ - K-FinHallu (Financial Verifiability): Gemini 2.5 Flash: 0.860 binary detection F1 (Paper Table 3, 28 May 2026), https://benchmarks.finance/leaderboards/k-finhallu/ - KTD-Fin (Financial trading and portfolios): Qwen3.6-Plus: 85.3 total return % (Paper Table 4, 27 May 2026), https://benchmarks.finance/leaderboards/ktd-fin/ - LiveTradeBench (Financial trading and portfolios): GPT-4.1: 6.3 cumulative return % (Paper Table 2, 5 Nov 2025), https://benchmarks.finance/leaderboards/livetradebench/ - Market-Bench (Financial trading and portfolios): grok-4: 443.2 mean MAE (marketbench.ai leaderboard, date not stated), https://benchmarks.finance/leaderboards/afterquery-market-bench/ - Moneda FilingsBench (Financial filings Q&A): Numa 1: 91.6 % correct (private split) (Moneda run, 2026-11-02, 2 Nov 2026), https://benchmarks.finance/leaderboards/moneda-filingsbench/ - Moneda PortfolioBench (Financial trading and portfolios): Numa 1: 96 Portfolio tasks, % correct (Moneda run, 2026-11-01, 1 Nov 2026), https://benchmarks.finance/leaderboards/moneda-portfoliobench/ - Open FinLLM Leaderboard (Finance knowledge): GPT-4o: 46.0 balanced score (mean of text/vision/audio averages) (Open FinLLM Leaderboard (scores.json), 16 Apr 2026), https://benchmarks.finance/leaderboards/open-finllm-leaderboard/ - PolyBench (Financial trading and portfolios): MiMo-V2-Flash: 17.6 Confidence-Weighted Return (CWR) % (Paper Table 5, 3 Apr 2026), https://benchmarks.finance/leaderboards/polybench/ - PortBench (Financial trading and portfolios): GLM-5.1: 0.384 CEPS (0-1) (Paper Table 2, 9 Oct 2026), https://benchmarks.finance/leaderboards/portbench/ - PRBench Finance (Financial advice and safety): Muse Spark 1.3: 59.5 weighted rubric score (0-100) (Scale Labs leaderboard, date not stated), https://benchmarks.finance/leaderboards/scale-prbench-finance/ - Reasoning Models Ace the CFA Exams (CFA mock-exam evaluation) (Finance knowledge): Gemini 3.0 Pro: 97.6 Level I accuracy % (Paper Table 5, 9 Dec 2025), https://benchmarks.finance/leaderboards/cfa-reasoning-models-ace/ - StockBench (Financial trading and portfolios): Qwen3-235B-Think: 2.5 final return % (Paper Table 2, 2 Mar 2026), https://benchmarks.finance/leaderboards/stockbench/ - Tax Agent Bench (Finance knowledge): Qwen 3.8 Max 0902: 49.7 accuracy % (all-pass x citation factor) (Vals.ai leaderboard, 29 Oct 2026), https://benchmarks.finance/leaderboards/vals-tax-agent-bench/ - Vals Finance Index (Analyst Financial research): Gemini 4 Carbon: 66.3 index accuracy % (Vals.ai leaderboard, 29 Oct 2026), https://benchmarks.finance/leaderboards/vals-finance-index/ Source: https://benchmarks.finance/best/ Created and maintained by Moneda.