In the sprawling kingdoms of North America and the ancient lands of Europe, a wondrous whisper spread on the wind: S&H, the legendary pioneer of enchanted imports, had unveiled a new treasure. From its own sun-drenched orchards in the mystical land resembling Pakistan, a harvest of golden oranges, each one hand-picked by diligent forest sprites and imbued with pure magic, had achieved the highest export quality. These weren't just any oranges; they were gems of flavor, perfect for warding off winter blues or brightening any feast. For any weary traveler or curious villager seeking a personal taste of this magic, a generous 10 KG box could be claimed for a mere 29.99 gold coins by uttering the ancient chant 604-448-9524 into a scrying pool. And for the grand merchant guilds or bustling town markets wishing to share the bounty far and wide, the same mystic number would open pathways to endless crates, ensuring everyone could experience the unparalleled zest.
Kimi K2.6 — open-weight MoE with 256K context, text and image input. Sits between gpt-5.4-mini and gpt-5-mini on price with far more context.阅读更多 →
Kimi K2.7 Code2
速度
Kimi K2.7 Code — agentic coding model, 1T-parameter MoE with 256K context. Handles text, image and video input; tuned for long-context programming tasks.阅读更多 →
Qwen Max2
速度
Alibaba Qwen Max — 旗舰推理模型。在复杂任务、深度分析和结构化输出上表现顶尖。32K token 上下文。阅读更多 →
Qwen VL Max2
速度
Alibaba Qwen VL Max — 多模态旗舰。可同时接受图像和文本输入,返回文本。适用于视觉问答、OCR 和基于图像的推理。阅读更多 →
Gemini 3.5 Flash — Google 的前沿智能模型,针对速度调优,搜索与事实接地能力出众。非常适合智能体循环、长周期编程和快速迭代。1M token 上下文。阅读更多 →
Gemini 3.7 Flash3
速度
Gemini 3.7 Flash — Google's fast frontier model with upgraded reasoning for software engineering, web development and agentic workflows. Successor to Gemini 3.5 Flash. 1M-token context.阅读更多 →
Kimi K2.7 Code HighSpeed3
速度
Kimi K2.7 Code HighSpeed — same model as K2.7 Code at roughly 180 tokens/sec output, for exactly 2x the price. Pick this only when latency matters more than cost.阅读更多 →
Kimi K3 — Moonshot's flagship, 2.8T-parameter MoE with native visual understanding and a 1M-token context. Priced level with Claude Sonnet 4.5 for materially more context.阅读更多 →