Deedy (@deedydas) on X

Deedy on X: "China just dropped a new model. ByteDance Doubao-1.5-pro matches GPT 4o benchmarks at 50x cheaper — $0.022/M cached input tokens, $0.11/M input, $0.275/M output — 5x cheaper than DeepSeek, >200x of o1 — 32k + 256k context — sparse MoE architecture AI truly too cheap to meter."

China just dropped a new model. ByteDance Doubao-1.5-pro matches GPT 4o benchmarks at 50x cheaper — $0.022/M cached input tokens, $0.11/M input, $0.275/M output — 5x cheaper than DeepSeek, >200x of o1 — 32k + 256k context — sparse MoE architecture AI truly too cheap to meter.
The above prices are for 32k, Doubao 1.5 pro 256k is priced at $1.23/M output. That is just above $1.1/M output of DeepSeek V3 (64k) but still far cheaper than $10/M of GPT 4o
You can actually access it on doubao.com with translation to English
These cheaper prices wouldn’t buy Sam Altman $2M cars.