Hacker News new | past | comments | ask | show | jobs | submit
It looks like it's being served extremely fast via Openrouter right now https://openrouter.ai/google/gemini-4-neon. 127 tokens per second. This is way faster than Carbon at 46 tokens per second (https://openrouter.ai/google/gemini-4-carbon). I'm a bit confused by this, since it seems like Neon is supposed to be their flagship model. Perhaps they've been throttling Carbon?

Carbon has a huge problem that it's just way too slow, and outputs way too much per task (on real world tasks, it's the worst at this https://micjm.com/blog/carbon-is-that-slow/).

Without the speed issue, Carbon would be a top model already. It's top on our battle system built on opencode, if you exclude early votes: https://agentduel.com/leaderboard?excludeEarlyVotes=true, though to be fair, small sample size so far.

Conclusion is that it's looking good for Neon. Even if it suffers from similar problems as Carbon, if it's even smarter and way faster, it should be a top model, maybe the best.

Edit: I was wrong about Neon being the larger model. I made this assumption because Neon has higher scores and they're both "v4".

Neon is an Opus sized model

Carbon is a Fable sized model

Carbon seems to be a version of Argon that is incredibly spendy with thinking tokens.

loading story #50552956
loading story #50552944
I wonder if this is an experiment with their new partial chip design that's vaguely similar to Taalas
loading story #50561291
loading story #50553021
Google has fully moved to rolling releases paradigm as their internal RSI starts to work, we'll see more rapid releases that improves on all fronts(size and capacity) and it starts to confuse the model size paradigm.

You can often have a new set of RL envs or RL algos that train a smaller model past the capacity of the bigger model, especially with distillation involved and the tradeoffs of RL rollouts vs pretrain.

> Carbon has a huge problem that it's just way too slow

And makes shit up, and costs significantly more than Sol/Opus on tasks requiring lots of decision making.