Importantly this restructuring does not negatively impact our full year 2023 guidance. Turning now to guidance. In Q4, we expect the same trends we have been seeing so far this year to continue. Namely, our new sales will be predominantly cloud and we expect our cloud ARR to grow by 35% year-over-year and because we are no longer selling a significant amount of self-managed subscriptions or perpetual licenses, self-managed subscription ARR and maintenance ARR is expected to decline on both a sequential and year-over-year basis. We delivered better than expected results again in Q3 and have good momentum going into Q4. That being said, Q4 is our biggest quarter of the year and still face a considerable amount of uncertainty in the macro environment.
Therefore, we believe it is prudent to reaffirm our previously issued full year guidance for revenue and ARR. With respect to our non-GAAP operating income and unlevered free cash flow, however, we are raising our full year guidance. We now expect non-GAAP operating income to be in the range of $430 million to $450 million, representing approximately a 25% year-over-year increase at the midpoint. We now expect adjusted unlevered free cash flow after tax to be in the range of $410 million to $430 million, representing approximately a 46% year-over-year increase at the midpoint. At this point in the year, our fourth quarter guidance is simply a derivative of our full year guidance, so I’ll just briefly give you the highlights. We expect GAAP total revenue will be in the range of $420 million to $440 million, representing approximately 8% year-over-year growth at the midpoint of the range.
We expect subscription ARR to be in the range of $1.098 billion to $1.118 billion, representing approximately 11% year-over-year growth at the midpoint of the range. We expect cloud subscription ARR to be in the range of $604 million to $614 million, representing approximately 35% year-over-year growth at the midpoint. And we expect non-GAAP operating income to be in the range of $130 million to $150 million, representing approximately 23% year-over-year growth at the midpoint. For modeling purposes, I’d like to provide a few more pieces of additional information. First, we expect adjusted unlevered free cash flow after tax for the fourth quarter to be in the range of $114 million to $134 million. Second, we estimate cash paid for interest will be approximately $40 million in the fourth quarter.
And for the full year, we estimate cash paid for interest will be approximately $149 million. Third, with respect to income taxes, our Q3 non-GAAP tax rate was 23% and we expect that rate to continue for the full year of 2023. We estimate full year 2023 cash taxes to be approximately $80 million. On a GAAP basis we expect the significant volatility of our income tax provision and rate to continue. For the full year 2023, we expect a GAAP tax provision in line with our cash taxes. And lastly our share count assumptions for the fourth quarter of 2023, we expect basic weighted average shares outstanding to be approximately 292 million shares and diluted weighted average shares outstanding to be approximately 297 million shares. For the full year 2023, we expect weighted average shares outstanding to be approximately $288 million and diluted weighted average shares outstanding to be approximately $293 million.
Yesterday, our Board of Directors approved a new share purchase authorization that enables us to buy up to $200 million of our Class A common stock through privately negotiated transactions with individual holders or in the open market. Our committee of the Board will determine the timing amount and terms of any repurchase. While we do not currently have any specific plans to purchase shares this authorization gives us the opportunity to move quickly if and when opportunities arise. And finally at our upcoming Investor Day on Tuesday December 5, we intend to provide a deeper understanding of Informatica’s business strategy, product innovation, growth drivers and financial objectives. Those interested in attending in person, please contact Victoria.
Operator, you can now open the line for questions.
Operator:
Operator: Thank you. [Operator Instructions] The question today comes from Alex Zukin with Wolfe Research. Your line is open.
Alex Zukin: Hey, guys. Thanks for taking my question. And congrats on a solid quarter. I guess, maybe just the first one. When you mentioned the macro conditions not really changing in the quarter. Can you maybe talk to what you’re seeing the combination of impacts we’ve heard obviously from all the hyperscalers at this point some of whom talk about attenuating optimization trends and solid AI workloads as being a driver for consumption. Can you maybe just talk about what you’re seeing out there in the market some of these larger projects that you’re attaching to? Because it does feel like you had some meaningful improvement in some of your consumption dynamics and cloud NRR actually went up. So I’m curious if you can tie that into any comment around stabilizing or bottoming trends there?
Amit Walia: Okay, Alex. Thanks for the question. Hope you are doing well? So look I think first is you have to really realize how unique we are. When you look at — when you ask the question of hyperscalers, we basically are the Switzerland of data. So when customers are looking to whether it’s an Azure or an AWS or GCP or in some cases a warehouse or a lake, we solve all of those use cases and master data management and governance use cases. So we kind of have a in-bid situation where we have basically sit in the middle of multiple different platforms. So when different demand can move from one to another it doesn’t really impact us 1:1. So that’s one. Second is, look, we’ve always been focused on mission-critical workloads.
And we’ve never chased what I’ve always said sometimes nice to have workloads that churn the fastest in a bad economy. And I think that has served us well. So when customers buy us we’re always buying us with the idea that I’m in it for the long haul. [indiscernible] opportunist. Very minimal on these nice to have workloads. And that strategy has always worked very well for us. Of course, having the best products and a single platform which also reduces the risk for our customer to go stitch together multiple products also helps them because they don’t have to spend more money and create more risk in stitching things together. And then last but not the least like what I’m hearing from customers a lot is, look customers are optimizing their spend data tied to AI is definitely amongst the top three if not the top two.
And all of those things are helping us. And lastly, like I said our partnerships we’ve been manically focused on that. We talked about the hyperscale, but the GSI partnerships with them building practices on us, building reference architectures and us all of those things create a, kind of, snowball effect or a tailwind effect as you might like to think about it Alex.
Alex Zukin: Perfect. And then maybe just as a follow-up. You mentioned the restructuring plan. Do you just need — because of the consumption dynamics that you’re seeing in your customer base? Do you just need less people to facilitate that? It’s more product-led roads and self-service? And then maybe just an update on our comment on Ithaca and the share repurchase authorization as well.
Amit Walia : I’ll take the first and Mike please cover the Ithaca one. So yes, and more to the first question that you said I think look I said that even the beginning of the year that as we transition to the final version of our transformation cloud only, there’s a lot of hybrid duplicative things paddle things that we were — we had that there was just no way around it. We took one round of simplification earlier this year, and certain other things were there and I’ve been saying that in every earnings call that there is more operating leverage that we feel we will be. And I think those are all the things that we are doing that are not necessarily needed as we think about the business. It’s a lot simpler for all the things you said and many other things we don’t need duplicative things to maintain or to do.
