PulseAugur
实时 19:26:23
English(EN) RT @eglyman: public benchmarks are saturated. every frontier model has trained against them, and the leaderboard tells you near nothing.

Fireworks AI推出Kimi K2.7 Fast,声称内部基准测试表现优异

Fireworks AI宣布推出Kimi K2.7 Fast,这是一个由Kimi K2.6 Fast技术驱动的更新的推理基础设施。该新产品可在Fireworks的无服务器平台上使用,提供标准(按token付费)和优先套餐选项,并即将推出快速通道。该服务支持256K上下文窗口,定价为每100万输入token 0.95美元,每100万输出token 4美元,每100万缓存命中0.19美元。此外,Fireworks AI认为,像Ramp开发的内部基准测试,比饱和的公共排行榜更有价值,可以评估AI模型。 AI

影响 通过大上下文窗口和分级服务选项增强AI应用的推理能力。

排序理由 这是推理基础设施服务的更新产品,而不是新的前沿模型发布。

在 X — Fireworks (inference infra) 阅读 →

AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →

报道来源 [2]

  1. X — Fireworks (inference infra) TIER_1 English(EN) · FireworksAI_HQ ·

    RT @richychn: the same tech that powered Kimi K2.6 Fast on Fire Pass now powers Kimi K2.7 Fast

    RT @richychn: the same tech that powered Kimi K2.6 Fast on Fire Pass now powers Kimi K2.7 Fast

  2. X — Fireworks (inference infra) TIER_1 English(EN) · FireworksAI_HQ ·

    RT @eglyman: public benchmarks are saturated. every frontier model has trained against them, and the leaderboard tells you near nothing.

    RT @eglyman: public benchmarks are saturated. every frontier model has trained against them, and the leaderboard tells you near nothing. w…