The transcriptions are slightly different, but they're both correct. You'll have to decide which style of transcription you prefer.
Phonetizer appears to be a more phonetic and less phonemic rendering; it marks long vowels, which is a little unusual for American English transcription -- it's not clear just what the vowel length signifies since it's not phonemic. It also distinguishes stressed [ʌ] and ustressed [ə], which are allophonic in American English. Upodn's outputs, by contrast, are closer to phonemic.
Both of them mark word stress, but there is no software that can reliably reproduce sentence stress, since that's not represented in English orthography. Word stress can be marked, because it's unchangeable for each word (a little parsing may be required to distinguish, say, record v. from record n.), but sentence stress is determined by semantics and pragmatics, not phonetics.
Neither of them offers Kenyon and Knott transcription, which has been the standard phonemic transcription of American English for 60 years or so, though several choices are offered.
Best Answer
I believe it's called "Quirk".
This was widely used in HTML (Quirks Mode) to maintain backwards compatibility with some old web pages: