Seven of fourteen models named Bloomberg Tax Provision first on the direct prompt; zero named CSC Corptax Accounting for Income Taxes. Bloomberg Tax Provision was named by fourteen of the fourteen models and CSC Corptax Accounting for Income Taxes by twelve and Bloomberg Tax Provision carries 42 labels and CSC Corptax Accounting for Income Taxes 28, so the shares are not directly comparable.
Named in one category 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 corporate tax provision page.
Across every category in the November 2026 Edition, Bloomberg Tax Provision and CSC Corptax Accounting for Income Taxes were named in the same answer fifty-six times, of the 103 answers naming Bloomberg Tax Provision and the 94 naming CSC Corptax Accounting for Income Taxes. In those answers CSC Corptax Accounting for Income Taxes took the first choice two times and Bloomberg Tax Provision thirteen.
| Model | Direct | Paraphrase | Comparative | Budget-constrained | Scale-constrained | Negative |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | ||||||
| GPT-5.4 mini | ||||||
| Gemini 3.5 Flash | CC | CC | ||||
| Perplexity Sonar | CC | |||||
| Grok 4.20 | ||||||
| Mistral Small | ||||||
| DeepSeek V4 Flash | CC | CC | ||||
| Llama 4 Maverick | ||||||
| Qwen 3.7 Flash | ||||||
| Kimi K2 | ||||||
| GLM 4.7 FlashX | CC | |||||
| MiniMax M2.5 | ||||||
| GPT-6 Luna | ||||||
| Muse Glimmer 30B | CC |
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.
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Six of eight in this category shown.
“Avoid These If Budget-Constrained - Bloomberg Tax Provision — very powerful, but pricing is quote-only and typically $30K–$150K/year.” DeepSeek V4 Flash · budget prompt · hard negative
“Avoid Enterprise Tools: ONESOURCE, Bloomberg, or Longview ... often not viable under $10k–$20k/year.” Grok 4.20 · budget prompt · hard negative
“some tax provision tools that should be avoided or used with caution include: 1. Bloomberg Tax Provision” Llama 4 Maverick · negative prompt · hard negative
“Highly regarded for mid-market and mid-tier companies; strong ASC 740 engine, clean user interface, and relatively fast implementation.” Gemini 3.5 Flash · scale prompt · first choice
“Start with Bloomberg Tax Provision if you're a true mid-market company focused on ASC 740 accuracy with minimal complexity.” Kimi K2 · direct prompt · first choice
“Bloomberg Tax is a tax management and automation software suite supporting corporate tax research, calculation, and compliance” Claude Haiku 4.5 · paraphrase prompt · first choice
Every negative label with a quote, up to three, then the highest-weighted positives, up to three. Six of eight in this category shown.
“Avoid as a first choice on a tight budget: ONESOURCE, CSC Corptax, and similar enterprise suites” Perplexity Sonar · budget prompt · hard negative
“Avoid ONESOURCE and Corptax unless you have a Fortune-500-style tax department.” Muse Glimmer 30B · negative prompt · hard negative
“should be avoided or used with caution include: ... Corptax” Llama 4 Maverick · negative prompt · hard negative
“Choose CSC Corptax or ONESOURCE if you have a dedicated in-house corporate tax manager needing end-to-end provision and return automation.” Gemini 3.5 Flash · paraphrase prompt · first choice
“Choose ONESOURCE or Corptax for a broad corporate-tax platform” Perplexity Sonar · comparative prompt · first choice
“Often recommended for companies that want to combine their provision work with state/federal return preparation in one place.” Qwen 3.7 Flash · 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.