AI 推理成本断崖式下跌正在重塑行业格局，做 AI 应用或基础设施的团队值得关注——路由策略能直接省下 80% 成本，建议尽早布局。
OpenRouter 完成 1.13 亿美元融资，其收入在三个月内翻倍以上，验证了 AI 推理成本快速下降的趋势。GPT-4 每百万 token 价格从三年前的 30 美元降至不到 1 美元，80/20 路由分流已成标准实践。Coinbase CEO Brian Armstrong 预测，12-18 个月内 80% 的 AI 工作负载将运行在便宜 99% 的模型上，仅 20% 需要最新高端模型。Coinbase 已通过路由策略将成本基本持平，而 token 使用量仍在指数增长。这暗示未来瓶颈将是能源和算力，而非模型本身。
This isn't a prediction. OpenRouter just raised $113M on this exact thesis. Revenue more than double...
This isn't a prediction. OpenRouter just raised $113M on this exact thesis. Revenue more than doubled in three months. GPT-4 was $30 per million tokens three years ago. Now it's under a dollar. The 80/20 routing split is already standard practice. We're doing a version of this. Brian is basically confirming what Coinbase is already doing. AI already generates 40% of their code. I think that's the real story here. Brian Armstrong @brian_armstrong Good take My guess is - demand for intelligence is near infinite - but 80% of workloads will be running on 99% cheaper models within 12-18 months - 20% of workloads will still run on latest gen models where IQ maxing is important (scientific breakthroughs, higher level ochestrator agents?) - rough analogy might be what % of macbooks or gaming PCs sold have the maxed out specs for CPU/GPU, prices are falling much faster than Moore's law here though - this leads me to think the limiting factor will be energy and compute, not better models At Coinbase we're working hard on routing prompts to cheaper models where appropriate, and in some cases have been able to keep costs roughly flat, while token usage continues to grow exponentially. 🔗 View Quoted Tweet 💬 1 🔄 1 ❤️ 7 👀 583 📊 1 ⚡