Welcome to our weekly discussion hub where we talk about different models and API platforms.
All general conversations about models and APIs should go here. Individual posts asking “which model is best” will be removed from the main forum.
Note: This thread is not for promoting your own services constantly. We might allow announcements for new legitimate services occasionally, but expect promotional content to be removed.
Thread Structure Guide
Look for these organized sections in the comments below:
- LARGE MODELS: 70B+ – Talk about models with 70 billion parameters and above
- MEDIUM-LARGE: 32B-69B – Discussion for models between 32B and 69B parameters
- MEDIUM: 16B-31B – Chat about models in the 16B to 31B parameter category
- SMALL-MEDIUM: 8B-15B – Share thoughts on models from 8B to 15B parameters
- SMALL MODELS: Under 8B – Discuss compact models below 8B parameters
- API SERVICES – Talk about API platforms, costs, speed, availability, etc.
- OTHER TOPICS – Everything else about models and APIs that doesn’t match above categories
Just find the right category and add your comment there. This way everyone can find what they need quickly and discussions stay neat.
Let’s get chatting!
Finally, a structured approach to these discussions. The constant “what’s the best model” posts were getting repetitive. One thing I’d suggest is maybe adding a section for multi-modal capabilities since that’s becoming increasingly important. Also worth noting that some providers are starting to offer fine-tuned versions of popular models, which can blur the lines between these parameter categories. I’ve noticed significant differences in how models handle specific domains like code generation versus creative writing, so it might be helpful if people mention their use case when making recommendations. The API section will definitely need frequent updates given how volatile pricing has been lately.
hey zack, appreciate you creating this space! it’s tough keeping up with all the models out there, especially when some get way too much hype. curious to see what everyone thinks in the different sections.
Good initiative setting up these weekly threads. The parameter-based categorization makes sense for organizing discussions, though I’d argue that parameter count doesn’t always correlate with performance anymore. Some of the newer 32B models I’ve tested outperform older 70B+ ones in specific tasks. The API services section will probably be the most active since pricing and reliability changes so frequently across providers. Looking forward to seeing actual benchmarks and real-world usage reports rather than just marketing claims.