adds new gpt-4o-2024-08-06 model

[jlvallelonga]

update for new "gpt-4o-2024-08-06" model and price

| Starting today, we're updating the default GPT-4o model to gpt-4o-2024-08-06, offering 50% lower input token costs and 33% lower output token costs! Point your code at the gpt-4o or gpt-4o-2024-08-06 model to get the savings.

[krschacht]

👍 looks good, just the one failing test to update then merge away