Open-model task cost: Muse uses fewer tokens than DeepSeek but costs three times more
What happened
On October 20, Computing Life compared the per-task cost of Muse Spark Open and DeepSeek. Artificial Analysis benchmarks put both at 39 points, but Muse Spark Open's output token count is less than half of DeepSeek's, while its average cost per task is $0.84 against $0.27, more than three times higher. The cost figure covers input, cache, reasoning chain-of-thought and final-answer tokens. Per-task cost depends on both token consumption and unit price, so output volume alone does not settle it.
Written by AI from the coverage · updated 2 hours ago
Coverage
Follow the reports to see the story from different sides.
- Computing Life · Share · Yage开放模型任务成本解析：为什么 Muse 比 DeepSeek 更省 token 却更贵？
开放模型的单任务成本同时受 token 消耗和单价影响，Muse Spark Open 输出量不足 DeepSeek 的一半，任务费用却高出三倍多。Artificial Analysis 的评测中，两者综合成绩均为 39 分，平均每任务费用分别为 0.84 美元和 0.27 美元，计算涵盖输入、缓存、推理思维链及最终答案的 token 费用。
Heat over time
Not enough continuous observations to draw a trend yet.