The one thing I love about Nixtla is their constant attention to the latest and greatest techniques that add value to forecasting projects. Because it's an open source implementation, it was easy to use and integrate into existing workflows. I primarily use Nixtla for almost all forecasting projects and can reach out easily to their support via community slack. Review collected by and hosted on G2.com.

In the past, Nixtla has been missing a few enterprise grade features that would have been nice to have at the time. Review collected by and hosted on G2.com.

Nixtla is a super innovative company with incredible customer focus, agility and state-of-the-art forecasting research experts. It was incredibly productive to work with Nixtla experts. Nixtla provided us access to their latest TimeGPT model and provided end-to-end hands on support on our complex fine-tuning use cases. TimeGPT model is very easy to use and very easy to onboard. We hosted TimeGPT on-premise with full control over our data and the model. It was very easy to integrate TimeGPT in our production environment. We used TimeGPT on a weekly frequency for our Supply Zone demand forecast. It was also very easy to implement TimeGPT for new business use cases and prototype quickly. Review collected by and hosted on G2.com.

There were no downside in our collaboration with Nixtla. Review collected by and hosted on G2.com.

What I like best about TimeGPT from Nixtla is how it helped me tackle a real challenge that I was struggling with: understanding why users were leaving for competitors and improving our demand and capacity forecasting to boost retention. Without the technical know-how or resources to build accurate forecasts myself, I needed a tool that could turn these insights into actionable predictions without requiring me to be a data scientist.

TimeGPT really stood out because it’s easy to set up, integrates effortlessly with our existing tools, and is quick to get running, so we could start using it right away without a long learning curve or disruptions. I use it regularly because it’s straightforward and fits smoothly into our daily workflows. On top of that, their customer support is always responsive and ready to help when I have questions.

I appreciate that Nixtla offers a powerful forecasting solution that strikes the right balance between technical depth and user-friendliness, making it possible for people like me to make smarter, data-driven decisions with confidence. Review collected by and hosted on G2.com.

The only minor downside I’ve noticed is that some of the documentation could be a bit more detailed to help less technical users get up to speed even faster. Review collected by and hosted on G2.com.

What I liked best about TimeGPT was its remarkable ease of implementation and fast prototyping capabilities. Setup was frictionless, allowing students to start making accurate forecasts within hours rather than days or weeks. The API integrates seamlessly with Python and Jupyter workflows, making it easy to incorporate into our existing curriculum. We used TimeGPT on a weekly basis throughout the semester, consistently benefiting from its fast, reliable results. This frequent use allowed us to focus on generating insights, evaluating models, and making informed decisions—without getting bogged down by complex traditional forecasting methods.

Thanks to TimeGPT, students graduate not only with theoretical knowledge but also with practical, production-ready forecasting skills that they can confidently apply in internships and real-world projects. The responsive customer support further ensured smooth adoption and quick problem resolution.

TimeGPT truly bridges the gap between academic learning and real-world application, preparing students to enter the workforce with a valuable, job-ready skill set. Review collected by and hosted on G2.com.

So far nothing ,everything has worked smoothly. Review collected by and hosted on G2.com.

Scalability, completeness, and open-source Review collected by and hosted on G2.com.

I don't see any downside in using Nixtla Review collected by and hosted on G2.com.

The Nixtlaverse uses a lot of best practices for software and ML that I love, with an easy-to-use API.

Examples include Conformal Prediction intervals, Probabilistic Adapters, Model Interpretability, and Horizontal/Distributed computing.

I also love the fact that MLForecast and NeuralForecast offer more than 30 SotA models.

And finally, I love the ease of use of TimeGPT. We are using it as one of the models in our proprietary library by self-hosting it in our infrastructure and have developed an API with Middleware to track its use and performance.

The Nixtla team has always been very helpful and responsive. Review collected by and hosted on G2.com.

I am being really picky with these, going for a 10 A+++ score, but:

https://github.com/microsoft/CBM remain unintegrated

Extending NeuralForecast is not the easiest.

SotA multivariate architectures scale poorly with the number of series. Not a Nixtla con per-se, but it has not been optimized. Review collected by and hosted on G2.com.

TimeGPT has completely changed the way our data science team approaches time series forecasting. We used to spend hours building and tuning individual models, which wasn’t scalable across hundreds of series. With TimeGPT, we can now generate accurate forecasts in seconds - it’s dramatically increased our experimentation speed and let us deploy faster.

The ease of integration, zero manual hyperparameter tuning, and consistent performance across different datasets made it a clear fit. It’s rare to find a tool that’s both powerful and effortless to implement.

The Nixtla team was also incredibly responsive - they partnered with us closely to align the tool with our use case in pharma forecasting. Overall, TimeGPT has become a critical piece in our pipeline. Review collected by and hosted on G2.com.

I wish i had a little more integration with other platforms, like Snowflake/Tableau, although I have heard that it recently was included in Snowflake, but would love to able to use it dashboards! Review collected by and hosted on G2.com.

TimeGPT is easy to integrate and delivers strong out-of-the-box performance for energy forecasting, anomaly detection, and demand prediction. It saves us time and helps us flag issues proactively. The Nixtla team has also been responsive and great to work with. Review collected by and hosted on G2.com.

Slightly more expensive compared to using your own tree-based models like lightgbm or even some other timeseries forecasting models on huggingface. Review collected by and hosted on G2.com.

Very easy to use, and fast. I was able to set it up quickly in my forecasting pipeline. Review collected by and hosted on G2.com.

It's bit of a black box. Not a problem if quick, reliable forecasts is all you want, but it's not clear what assumptions are being made. Review collected by and hosted on G2.com.

Personally and through my organizations, we have used variants of Nixtla's offerings for probably 10+ forecasting use cases. A couple of things come to mind: the Nixtlaverse is comprehensive, covering a wide range of features and tools needed from classical to foundation models and switching between them easily. This allowed us to go to production fast and easily and was, compared to other tools that we have used, a factor of 5 faster.

Teams use Nixtla daily at Zalando and weekly at Databricks. Review collected by and hosted on G2.com.

Nixtla requires some work to build a vertical end-to-end solution. They don't claim this and it's hard to build a generic solution for more vertical integration. Review collected by and hosted on G2.com.