# Mercury vs Bank of America: which do AI models recommend for business banking, September 2026

Finance AI Recommendation Index, September 2026 Edition, Business banking. One of twelve models named Mercury first on the direct prompt; one named Bank of America. Page: https://finance-ai-index.com/treasury/business-banking/mercury-vs-bank-of-america/

| | First-choice share | Rank | Negative rate | Labels | Models naming it |
|---|---|---|---|---|---|
| Mercury | 20% | #1 of 13 | 7% | 43 | 12 of 12 |
| Bank of America | 2% | #8 of 13 | 22% | 27 | 12 of 12 |

## The direct prompt, model by model

- Kimi K2: mercury first (first choices: Chase Business Banking, Mercury) (alternatives: Brex, First Internet Bank)
- Grok 4.1 Fast: bank of america first (first choices: Bank of America, Chase Business Banking) (alternatives: Brex, PNC Bank, Rho)
- GPT-5.4 mini: neither first, one named (first choices: Chase Business Banking) (alternatives: Bank of America, TD Bank, U.S. Bank)
- Gemini 3.5 Flash: neither first, one named (first choices: Rho) (alternatives: Bank of America, Brex, Chase Business Banking, Fifth Third Bank, Mercury, PNC Bank)
- DeepSeek V4 Flash: neither first, one named (first choices: Brex) (alternatives: Bank of America, Chase Business Banking, HSBC, Mercury, Relay, U.S. Bank, Wells Fargo)
- MiniMax M2.5: neither first, one named (first choices: Chase Business Banking) (alternatives: Airwallex, Bank of America, Barclays, HSBC, Mercury, NatWest, Rho, Silicon Valley Bank, Wells Fargo)
- Claude Haiku 4.5: neither named (first choices: Brex) (alternatives: Airwallex, Chase Business Banking, U.S. Bank)
- Perplexity Sonar: neither named (first choices: Rho) (alternatives: Bluevine, Brex, Relay)
- Mistral Small: neither named (first choices: Axos Bank, PNC Bank) (alternatives: First Internet Bank, Revolut Business, Wise Business)
- Llama 4 Maverick: neither named
- Qwen 3.7 Flash: neither named (first choices: Bluevine) (alternatives: Airwallex, Chase Business Banking, Relay)
- GLM 4.7 FlashX: neither named (first choices: Chase Business Banking) (alternatives: Airwallex, Qonto, Wise Business)

## What the models said about Mercury

- "Frequent account freezes/closure, slow support, funds held 60+ days | 15.4 negative reviews/month (4x industry average)" (DeepSeek V4 Flash, negative prompt, hard negative)
- "neobanks (such as Mercury, Bluevine, or Novo) rely heavily on automated fraud-detection algorithms" (Gemini 3.5 Flash, negative prompt, soft negative)
- "does not support physical branches or cash deposits" (Perplexity Sonar, budget prompt, soft negative)
- "I'd usually recommend Mercury if your priority is a modern, startup-friendly banking platform" (GPT-5.4 mini, paraphrase prompt, first choice)
- "For most mid-market B2B companies today, Mercury or Chase are the safest starting points" (Kimi K2, direct prompt, first choice)
- "Mercury, which is considered the best overall digital business banking platform for startups" (Llama 4 Maverick, paraphrase prompt, first choice)

## What the models said about Bank of America

- "Bank of America has the most complaints of any bank in the U.S., with customers reporting poor customer service, long wait times, excessive fees" (Claude Haiku 4.5, negative prompt, hard negative)
- "128k+ complaints, A- BBB but high unresolved issues (e.g., frozen funds, payment delays, account closures)" (Grok 4.1 Fast, negative prompt, hard negative)
- "Relationship banking, SME/mid-market leader ... Recent Greenwich surveys rank Chase/BofA/JPM high for mid-market trust/service." (Grok 4.1 Fast, direct prompt, first choice)
- "stick to Chase or Bank of America" (Gemini 3.5 Flash, comparative prompt, first choice)
- "Best if you need broad branch access and a familiar commercial platform: Bank of America Cash Management / CashPro" (GPT-5.4 mini, direct prompt, alternative)

Share is the count of first choices across the direct, paraphrase, budget and scale prompts over all twelve models, for a mid-market B2B company; rank is within the category. Comparisons are drawn for the top eight products in each category. Published under CC BY 4.0; the output is the models' output, and nothing here is a recommendation by the index.