The closest thing to a standard for academic pronunciation of Attic Greek is described in Allen’s Vox Graeca. Schools in the UK have been following a “scientific pronunciation” that more or less conforms to it since the end of the 19th century. This replaced an 18th-century pronunciation scheme called “Erasmian” which was was reasonably accurate in Erasmus’ day, but lost accuracy in English-speaking countries due to the great English vowel shift.
(Here’s an American explaining the great English vowel shift with a Texan accent: https://www.youtube.com/watch?v=UOfYck3J3e0)
A copy of Vox Graeca (it’s written in English despite the Latin title) and the resources at forvo.com are a good start to Greek pronunciation. But make sure that you listen to British BBC pronunciations, otherwise you’ll get into trouble with ω and ο! We Americans pronounce “father” and “saw” differently, and Allen uses these as demonstration words.
Randall Buth has been a big proponent of “spoken Koine” in recent years. This pronunciation is a modern Greek pronunciation with some vowel quality modifications, taken from Horrocks’ description of 4th century A.D. Koine.
And then there is modern Greek, used by many. The unfortunate thing about modern Greek is that the dozens of vowel quality distinctions maintained by the ancients (of both length and quality), have collapsed down to 8 in modern Greek. Length distinction has completely disappeared and many vowels sound like word “E” from the English phrase “the letter E.” This is known as iotization.
The quality collapse in modern Greek makes it impossible to tell important words like “you” (ὑμεῖς) from “us” (ἡμεῖς). The quantity collapse makes it impossible to read Greek poetry. On the other hand, many people have no trouble with using the Modern or Buth/Horrocks pronunciations for reading the Bible or later Greek Patristic writers.
The pitch accent is a more complicated subject. Pitch accents are difficult. You can look up videos of foreigners speaking Japanese or Norwegian to verify this to yourself. Also, properly reconstructing a pitch-accent is impossible. While we know (through a glass darkly) the word-level pitch properties of ancient Greek, we don’t know the phrase-level properties. Classicists tend to either use stress like the modern Greeks do (and in fact Allen suggests doing this) or make bad attempts at pitch. Personally, though I am no classicist, I make bad attempts at pitch. The stress accent, I have found, tends to interfere with poetry.
Diphthongs are complicated by the presence of glides, and it’s worth reading the chapters on Allen for information here. The only real question in long diphthongs. Allen suggests to drop them entirely, as Greek did sometime after the second century B.C., but many of us attempt them on various schemes.
The following video is my own attempt at the alphabet. (This is after about 4 years of Greek study and reading Greek aloud, so it represents a beginner’s attempt to give you an idea rather than anything authoritative):
https://youtu.be/mm2MNIJ5zd4
The alphabet video is iambic verse, each line a repetition of three “di-dah di-dah” segments. In English poetry we do “unstressed-stressed” for the “di-dah” metron, but the Greeks did “short syllable-long syllable.” The pitch accent (which I attempt, but badly, as mentioned before) would matter for the tune if it were sung, but does not affect the rhythm of the verse.