Assuming 100% of that raise goes to LLM costs (almost certainly unrealistic because of needs to pay salaries, etc), and assuming a conservative $100/month in frontier model LLM charges when doing serious coding, that raise amounts to $5.7M / $100 = 57,000 user-months, or less than 5,000 users for a year.
It's hard to imagine a scenario where this raise isn't completely burned through in a couple of months, tops.
charv•1h ago
Not quite sure how I feel about that, but I guess the major AI players are already using chats for model training anyways. Reminds me a bit of OpenRouter with grok code fast 1, trading off hot models for usage stats.