Anrok vs Numeral
One of fourteen models named Anrok first on the direct prompt; zero named Numeral. Anrok was named by twelve of the fourteen models and Numeral by ten and Anrok carries 27 labels and Numeral 17, so the shares are not directly comparable.
Named in three categories this edition.
Named in one category this edition.
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; every quote names the model and the prompt it came from. Both figures come from the sales tax automation page.
By framing
AnrokFirst choices, of fourteen modelsNumeral
Across every category in the November 2026 Edition, Anrok and Numeral were named in the same answer twenty-four times, of the 88 answers naming Anrok and the 47 naming Numeral. In those answers Numeral took the first choice zero times and Anrok four.
Every model, every framing
| Model | Direct | Paraphrase | Comparative | Budget-constrained | Scale-constrained | Negative |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | | | ||||
| GPT-5.4 mini | | | ||||
| Gemini 3.5 Flash | | | | | | |
| Perplexity Sonar | | |||||
| Grok 4.20 | | | | |||
| Mistral Small | | | ||||
| DeepSeek V4 Flash | | |||||
| Llama 4 Maverick | | | ||||
| Qwen 3.7 Flash | | |||||
| Kimi K2 | | | | |||
| GLM 4.7 FlashX | | |||||
| MiniMax M2.5 | | | | |||
| GPT-6 Luna | ||||||
| Muse Glimmer 30B | | | |
Anrok Numeral first choice named as an alternative argued againstblank: not namedEach cell is one answer, Anrok on the left and Numeral on the right.The direct prompt
Anrok first, Numeral not the choice
1 of 14 modelsNumeral was named in the answer but not as the choice, or not at all.Neither was the first choice, one was named
6 of 14 modelsThe answer put something else first and named one of the two as an alternative.Neither was named
7 of 14 modelsThe answer made no first choice from these two in this category.Bold names in an answer are the products the judge labeled a first choice; a model naming several gives each of them that label. The full answer text for every row is in the record.
By buyer segment
Anrok0%#7 of 10Numeral0%#– of 10What the models said about Anrok
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Five of five in this category shown.
“Grouped in warnings about assuming full automation. It can mishandle edge cases like digital products, international sales” Grok 4.20 · negative prompt · hard negative
“the platforms with the most charges that exist without a published number are Avalara, Anrok, TaxCloud, Zamp” Muse Glimmer 30B · negative prompt · soft negative
“B2B SaaS / Software Company | Anrok | Solves subscription changes, prorations, and complex SaaS state tax laws.” Gemini 3.5 Flash · comparative prompt · first choice
“Go with Anrok if you are a pure-play B2B SaaS / digital product business” Gemini 3.5 Flash · paraphrase prompt · first choice
“If you're a software/SaaS company with modern billing → Anrok” Kimi K2 · direct prompt · first choice
What the models said about Numeral
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Five of six in this category shown.
“Numeral's reported filing reliability issues are a concern for businesses with real compliance obligations” Claude Haiku 4.5 · negative prompt · soft negative
“Quote-only | Usage overages” Kimi K2 · negative prompt · soft negative
“Numeral is often the best sales tax solution for ecommerce and SaaS companies, wholesalers, and manufacturers.” Llama 4 Maverick · budget prompt · first choice
“Excellent for growing SaaS/ecommerce or simpler needs. AI-powered calculation, filing, and expert review at potentially lower cost.” Grok 4.20 · direct prompt · alternative
“Numeral is a true full-service solution for DTC and marketplace sellers, subscription software, and B2B sellers” Claude Haiku 4.5 · direct prompt · alternative
Comparisons are drawn for the top eight products in each category, each against each. The output is the models' output; nothing here is a recommendation by the index.