# Melissa vs Google Address Validation API: which do AI models recommend for address verification, October 2026

GTM AI Recommendation Index, October 2026 Edition, Address verification tools. Zero of fourteen models named Melissa first on the direct prompt; zero named Google Address Validation API. Page: https://gtm-ai-index.com/gtm-data/address-verification/melissa-vs-google-address-validation-api/

| | First-choice share | Rank | Negative rate | Labels | Models naming it |
|---|---|---|---|---|---|
| Melissa | 8% | #3 of 7 | 5% | 39 | 14 of 14 |
| Google Address Validation API | 4% | #5 of 7 | 10% | 10 | 8 of 14 |

## The direct prompt, model by model

- Gemini 3.5 Flash: neither first, one named (first choices: Loqate) (alternatives: Lob, Melissa, PostGrid Address Verification, Smarty)
- Perplexity Sonar: neither first, one named (first choices: Smarty) (alternatives: Loqate, Melissa, PostGrid Address Verification)
- Grok 4.1 Fast: neither first, one named (first choices: Smarty) (alternatives: Melissa, PostGrid Address Verification)
- DeepSeek V4 Flash: neither first, one named (first choices: Smarty) (alternatives: Loqate, Melissa, PostGrid Address Verification)
- Qwen 3.7 Flash: neither first, one named (first choices: Smarty) (alternatives: Loqate, Melissa)
- GLM 4.7 FlashX: neither first, one named (first choices: Smarty) (alternatives: Experian Address Validation, Melissa, PostGrid Address Verification)
- MiniMax M2.5: neither first, one named (first choices: Smarty) (alternatives: Melissa, PostGrid Address Verification)
- GPT-6 Luna: neither first, one named (first choices: Smarty) (alternatives: Lob, Loqate, Melissa)
- Claude Haiku 4.5: neither named (first choices: PostGrid Address Verification) (alternatives: Informatica Data as a Service, Melissa Address Verification, Smarty)
- GPT-5.4 mini: neither named (first choices: Melissa Global Address Verification) (alternatives: Experian Address Verification, PostGrid Address Verification, Smarty)
- Mistral Small: neither named (first choices: Experian Address Validation, Melissa Global Address Verification) (alternatives: Loqate, PostGrid Address Verification)
- Llama 4 Maverick: neither named
- Kimi K2: neither named (first choices: Smarty) (alternatives: Loqate, Melissa Global Address Verification)
- Muse Glimmer 30B: neither named (first choices: Loqate by gbg, Smarty) (alternatives: Melissa Address Verification)

## What the models said about Melissa

- "General cautions for all address verification tools (e.g., Smarty, Melissa, Quadient, Shufti Pro, SumSub, Veriff, DataFlow)" (Grok 4.1 Fast, negative prompt, soft negative)
- "Melissa is generally heavier and more complex to implement. It is less of an "add-on" tool for a website" (Qwen 3.7 Flash, comparative prompt, soft negative)
- "Melissa - Offers a comprehensive address verification API with features such as real-time address validation, standardization, and enrichment." (Llama 4 Maverick, paraphrase prompt, first choice)
- "The Industry Standards (High accuracy, widely used): Smarty (formerly SmartyStreets), Loqate (Onbuka), Melissa (DataBox)." (Qwen 3.7 Flash, scale prompt, first choice)
- "Melissa offers the best free tier for small businesses needing additional validation features" (MiniMax M2.5, budget prompt, first choice)

## What the models said about Google Address Validation API

- "Generally not CASS-certified, so less ideal for postal compliance needs" (DeepSeek V4 Flash, comparative prompt, soft negative)
- "I'd typically recommend starting with Google Address Validation or Melissa" (MiniMax M2.5, paraphrase prompt, first choice)
- "1. Google Address Validation API — Best Free Ongoing Option" (Kimi K2, budget prompt, first choice)
- "U.S. ecommerce checkout: Compare Smarty and Google" (GPT-6 Luna, comparative prompt, first choice)

Share is the count of first choices across the direct, paraphrase, budget and scale prompts over all fourteen 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.