Abstracts can be tricky to write because they are so succinct so every word matters. The most difficult part is figuring out the best way to frame your specific scientific question within the larger framework of the problem in the field that you are trying to solve. I find it helpful to outline my abstracts before I write them.
Here's an outline that I have found helpful when writing and editing my own abstracts:
What is the big problem in the field?
What is the specific problem I'm trying to answer?
What is stopping the field from answering that specific question?
How can I solve this with my particular method/model (ie What is special about this model)?
What did I do?
What did I find?
How did what I find help solve the specific problem in the field?
Why is this important to the big problem in the field?