Another arrival in the GenViz stable: Claude Sonnet 5.5. You can now choose Anthropic → Sonnet 5.5 in the model picker and put it to work on your business data, using the connections and analysis tools already in your conversations.
It joins Opus 5.5 and our other models as another option for conversation analytics: ask a question, inspect the result, and follow up while the details are still fresh.
A new option for everyday questions
In its Sonnet 5.5 announcement, Anthropic describes a faster model for well-scoped everyday tasks. It reports output generation more than 30% faster than Sonnet 5, with the same per-token API pricing and fewer tokens needed for many tasks.
That makes routine analysis a useful place to try it in GenViz: a grouped total, a comparison between periods, or a closer look at the rows behind an unexpected number. The speed figures are Anthropic’s measurements; response time in GenViz also depends on your queries, data, and analysis steps.
Start with a question you can check
Take a familiar question from your weekly review:
Show sales for the last four complete weeks, grouped by product category. Exclude cancelled orders and show the weekly totals in a table.
Open the returned table and inspect the query, including the dates and filters. Then ask a focused follow-up:
Which category had the largest decline in the most recent week? Show the five products that contributed most to that change.
From there, ask for a chart of the weekly trend or the underlying rows for a particular product. Each result stays part of the conversation, so you can move between the question, the calculation, and the evidence.
Saved extracts keep tables available for later inspection. For calculations across those results, Python analysis can return another table or chart with code you can review.
Choose how much reasoning to use
Sonnet 5.5 starts at high reasoning effort in GenViz, matching Anthropic’s documented default. The picker offers low, medium, high, xhigh, and max. Try a lower setting for a straightforward lookup, or increase it when a question needs more careful work across several steps.
Keep the question and expected answer in view as you compare settings or models. Check that the result uses the right business definition, accounts for missing values, and supports its conclusion with data you can inspect.
Try Sonnet 5.5 in your next conversation
Open the model picker and choose Anthropic → Sonnet 5.5. Availability follows your organization’s Anthropic access. On the Team plan, an administrator needs to configure an Anthropic API key with access to the model.
Sonnet 5.5 uses the same read-only database tools and connection controls as the other models in GenViz. Your prompts and the context used to answer them, including relevant schema information and query results, are sent to Anthropic. Our privacy policy explains the services involved.
Bring a question whose answer you know, check the result, and then ask the question you’ve been meaning to investigate.
Explore conversation analytics or download GenViz to get started.