[Feature] Increase GPT-4o-mini and Claude 3.5 Sonnet Token Limit

[EnzoAndree]

Problem Description The current token limits for GPT-4o-mini (8k) and Claude 3.5 Sonnet (4k) restrict the ability to produce longer outputs.

Proposed Solution

• Increase the output token limit for GPT-4o-mini from 8k to 16k, as mentioned in the recent OpenAI update (https://openai.com/index/gpt-4o-mini-advancing-cost-efficient-intelligence).
• Increase the output token limit for Claude 3.5 Sonnet from 4k to 8k (https://docs.anthropic.com/en/docs/about-claude/models#model-comparison).