Skip to main content
2diff.io

AI Models and Providers

BizInsights and BizAnalyzer draw on the same curated catalog of models from three providers, and your plan decides which of them you can select. Each model brings different strengths. Switch anytime. No vendor lock-in.

How AI works in BizAnalyzer

Select Your Model

Pick from a curated set of models across OpenAI, Anthropic, and Google directly in the chat interface. New models are added as they ship. Switch between models mid-conversation if you want a second opinion.

Same Data, Different Lens

Every model sees the same financial statements, ratios, and trends. Different models may emphasize different aspects of your data or present insights in different ways.

No Vendor Lock-In

If one provider has an outage or changes their pricing, your analysis continues with another. Your data and conversation history are not tied to any single AI vendor.

Transparent by Design

You always see which model is responding. No black box. No hidden switches. You know exactly what AI is analyzing your financial data.

Available Models

The starting model depends on your plan: BizInsights opens on Claude Haiku 4.5, and BizAnalyzer opens on Claude Opus 4.7. You can switch to any model your plan allows at any time.

OpenAI

OpenAI provides reasoning models across the GPT-5 and GPT-5.5 families, from premium multi-step analysis down to lightweight budget options.

ModelTierPlanContextBest for
GPT-5.5PremiumStandard128KNewest GPT reasoning model for coding and professional work
GPT-5PremiumStandard128KMost capable, complex reasoning
GPT-5 MiniStandardBizInsights128KBest balance of quality and cost
GPT-5 NanoBudgetFree demo32KUltra-cheap fallback

The Premium GPT models give the deepest reasoning; the Mini tiers offer solid everyday reasoning at lower cost; the Nano tiers are the cheapest reasoning-capable options for simple questions.

Anthropic (Claude)

Anthropic's Claude models are known for careful, nuanced analysis and strong performance on financial interpretation tasks.

ModelTierPlanContextBest for
Claude Opus 4.8PremiumStandard500KNewest Opus, most capable reasoning and agentic coding
Claude Opus 4.7DefaultPremiumBizInsights200KLatest Opus, top-tier reasoning and analysis
Claude Sonnet 4.6StandardBizInsights200KBest balance of intelligence and speed
Claude Haiku 4.5BudgetFree demo200KFast and cost-effective

Claude Opus 4.7 is BizAnalyzer's starting model. It handles the full range of financial analysis questions well, with structured explanations, ratio interpretation, trend analysis, and plain-language summaries of financial statements. For the highest reasoning depth, Claude Opus 4.8 is the newest top-tier option. For the fastest responses on simple lookups, Claude Haiku 4.5 or GPT-5 Nano are the lightest options.

Google (Gemini)

Google's Gemini models offer massive context windows, up to 1 million tokens, which means they can process extensive financial histories in a single analysis.

ModelTierPlanContextBest for
Gemini 2.5 ProPremiumStandard1MMost capable Gemini, complex reasoning
Gemini 2.5 FlashStandardBizInsights1MBest price and performance with thinking
Gemini 2.5 Flash-LiteBudgetFree demo1MFastest, lowest cost Gemini

The 1 million token context window is particularly useful if you have years of financial history. Gemini can hold all of it in memory at once, making it effective for long-range trend analysis.

Access and AI Request Limits

Where you can use the AI, which models you get, and the request limits on each surface. The public demo serves a small two-model allowlist and is rate-limited per visitor plus capped globally; signed-in subscribers run on the full AI Insights endpoint with anti-abuse rate limits, not per-question metering.

WhereModelsAI request limits
Try Demo (no login)
/apps/bizanalyzer/demo and /apps/bizinsights/demo
2 models (Claude Haiku 4.5, GPT-5 Mini)6 requests/min per visitor, 60/day per visitor, 2M tokens/day across all demo traffic
BizInsights ($19.99/mo)
/apps/bizinsights
7 models (includes the default Claude Opus 4.7 plus every budget model)120 requests/min, 1000/hour, 5000/day
BizAnalyzer ($29.99/mo)
/apps/bizanalyzer
11 models (everything in BizInsights, plus GPT-5.5, GPT-5, Claude Opus 4.8, Gemini 2.5 Pro)120 requests/min, 1000/hour, 5000/day

Public-demo traffic is also subject to a global 2-million-token-per-day ceiling shared across all visitors; the per-visitor caps above bound any single IP. Signed-in subscribers run on the full AI Insights endpoint and are anti-abuse rate-limited (the numbers above), not metered per question.

Which Models Come With Each Plan

The same model gating as the catalog above, expressed plan by plan: what you get on every plan, and what each paid tier adds on top. Every plan includes everything in the plans below it, and you can switch between your available models at any time during a conversation.

On every plan

3 models

Fast, everyday models included with every signed-in account.

ModelProviderBest for
GPT-5 NanoOpenAIQuick lookups, simple calculations, budget-friendly
Claude Haiku 4.5Try free in demoAnthropic (Claude)Fast responses, good for quick financial questions
Gemini 2.5 Flash-LiteGoogle (Gemini)Fastest responses, lowest cost, quick lookups

BizInsights ($19.99/mo) adds

4 models

Premium reasoning for deeper financial analysis.

ModelProviderBest for
GPT-5 MiniTry free in demoOpenAIEveryday financial questions, solid reasoning at lower cost
Claude Opus 4.7DefaultAnthropic (Claude)Latest Opus, top-tier reasoning and deep financial analysis
Claude Sonnet 4.6Anthropic (Claude)Best balance of intelligence and speed, strong at explaining financial concepts
Gemini 2.5 FlashGoogle (Gemini)Fast analysis with built-in reasoning, good all-around choice

BizAnalyzer ($29.99/mo) adds

4 models

The newest flagship models from each provider, on top of everything in BizInsights.

ModelProviderBest for
GPT-5.5OpenAIOpenAI's newest reasoning model for complex, professional analysis
GPT-5OpenAIComplex reasoning, nuanced financial analysis, multi-step questions
Claude Opus 4.8Anthropic (Claude)The newest Opus, the most capable reasoning available
Gemini 2.5 ProGoogle (Gemini)Deep analysis, complex multi-year comparisons

How to Choose

If you are not sure which model to use, start with your plan’s default (Claude Haiku 4.5 on BizInsights, Claude Opus 4.7 on BizAnalyzer).

When to try a different model:

  • Complex multi-step analysis: GPT-5 or Claude Opus 4.7 for the deepest reasoning
  • Quick factual lookups: GPT-5 Nano or Gemini 2.5 Flash-Lite for speed
  • Detailed explanations: Claude Sonnet 4.6 for step-by-step financial concept breakdowns
  • Large data context: Gemini 2.5 Pro for its 1M-token context window across years of history
  • Second opinion: Switch models to see how a different AI interprets the same data

You can change models at any point during a conversation. Your chat history is preserved when you switch.

Privacy and Data Handling

Financial data requires the highest level of protection. Here is how BizAnalyzer handles AI and your data:

  • Not used for training: Your financial data is never used to train AI models. It is processed for your query and discarded.
  • Not retained: AI providers do not store your financial data after generating a response.
  • Not shared: Your data is never shared with other users, third parties, or advertisers.
  • Minimal data sent: Only the financial data relevant to your question is included in the AI request. We do not send your entire accounting history for every query.
  • Encrypted in transit: All communication with AI providers uses TLS 1.3 encryption.

For complete details on our security practices, see our Security and Data Protection page.

See it working on your data.

BizAnalyzer connects to your accounting software via Intuit's official API. BizInsights runs on the figures you enter yourself. Either way, built on Google Cloud Platform infrastructure, payments by Stripe.

This content is for educational purposes only and does not constitute financial, tax, accounting, or legal advice. Consult with a qualified professional for advice specific to your situation. AI-generated insights may contain errors; verify information before taking action.Intuit and QuickBooks® are registered trademarks of Intuit Inc. Used with permission. Second Difference Solutions®, LLC is an independent software provider and is not affiliated with, endorsed by, or sponsored by Intuit Inc. Our applications integrate with QuickBooks® products via Intuit's official APIs.

OpenAI, GPT, Claude, Anthropic, Google, and Gemini are trademarks of their respective owners.