- Introducing GPT-4 Turbo 128K: Now live on our platform supporting contexts up to 96,000 words. It’s the new front-runner for context capacity, beating Claude V2’s 80,000 words and comes at 8 coins per 100 words, making it about 60% more cost-effective than GPT-4 8K.
Model Enhancements:
- Default Free Model Upgraded: GPT-3.5 Turbo 16K replaces the 4K as the new standard for our free default model.
- Google Palm 2 Boosted: Context expanded from 8K to 32K.
- Mythomax Updated: Upgraded context from 4K to 8K.
Direct Sourcing and Performance:
- We’ve streamlined our AI models. GPT models now come directly from OpenAI, enhancing response speeds noticeably. OpenRouter continues to supply our other models.
Despite the “Preview” label, we are confident that GPT-4 Turbo 128K is ready for heavy use. We encourage you to integrate it fully into your workflows.
We look forward to your feedback
