Bleanx 1.1

What’s new

Bleanx 1.1 brings a refreshed lineup of AI models from three providers. Every model works with every section, and the choice is always yours: you decide which model to use, and in which section.

Light models — recommended

• Claude Haiku 5.5 — Anthropic
• GPT 6 Luna — OpenAI
• Gemini 3.5 Flash Lite — Google

Middle models

• Claude Sonnet 5.5 — Anthropic
• GPT 5.6 Terra — OpenAI
• Gemini 3.8 Flash — Google

Why we recommend light models

The quality of an AI draft depends less on the size of the model and more on what it receives. That is where Bleanx does the heavy lifting.

Before any model writes a word, Bleanx gathers the context that section needs: your Business Core Concept, the related sections you’ve already saved, and the structure the section must follow — even when that context is spread across many connected parts of your plan. The model doesn’t have to guess what your business is. It’s given what it needs to know.

With context this complete, light models handle the work very well. They’re fast, consistent and use the least of your AI allowance. That’s why we recommend them for everyday drafting.

When a middle model is worth it

Middle models can write with more nuance and depth. In our experience, for most sections the result is very close to what a light model produces — while a middle model can use roughly five to ten times more of your allowance for the same section.

There are moments when that difference matters: sections that draw on an unusually large amount of connected context, or summaries that bring together many parts of your plan. If you want a more considered draft for a specific section, switch to a middle model for that one task, and switch back afterwards.

Why there are no flagship models

You won’t find the largest, most expensive models in Bleanx — and that’s a deliberate choice. Because our engine gives every model a focused, well-structured context, we haven’t found the extra cost of a flagship model worth it for business planning. We’d rather your allowance go further. If you ever hit a case where you need more, tell us at support@bleanx.org.

Choosing a provider is choosing a style

Anthropic, OpenAI and Google models all receive the same context about your business. Where they differ is style: how they phrase, structure and emphasize. Try a section with each and keep the style that reads most like your business.

What stays the same

Your AI allowance, Usage Credits and permissions work exactly as before. Formulas still calculate every financial total, and AI model providers don’t train on your data.