长上下文榜
智客 ZICQ 大模型榜单每日更新,覆盖智能、编程、价格与上下文,便于选型与比价。
最近更新:45 分钟前(2026-11-19 05:12)数据来源与评测方法
按 context_length 倒序 —— 适合长文档阅读、整库分析、百万级 token 任务。
2. 评测基准:智能、编码与智能体指数直接源自第三方国际权威评测平台 Artificial Analysis (artificialanalysis.ai) 官方指标。
3. 综合榜权重:综合智能分 = 60% × 综合智力指数 + 25% × 代码生成指数 + 15% × 智能体与工具调用指数。无评测基准数据的模型排在综合榜末尾。
4. 推广链接:标注「推广」的「去调用」按钮是推广链接,ZICQ 可能因此获得佣金;推广关系不影响任何排名与评分。
当前筛选的统计
⚡ 模型并排 PK 对比器 & Token 成本计算器 点击下方任意模型的「+ PK 对比」即可替换 · 实时测算月度 API 成本 · 支持一键复制对比链接
| 对比维度 | xAI: Grok 4.20 Multi-Agent xAI | xAI: Grok 4.20 xAI | Auto Router Openrouter |
|---|---|---|---|
| 综合智能 / 编码 / 智能体 | — | — | — |
| 上下文窗口 | 2M tokens | 2M tokens | 2M tokens |
| 输入 / 输出单价 (1M) | $1.25 / $2.50 | $1.25 / $2.50 | $-1000000.00 / $-1000000.00 |
| 预估月度账单 (30天) | 输入参数自动测算 | 输入参数自动测算 | 输入参数自动测算 |
-
#301
Qwen3.6-35B-A3B is an open-weight multimodal model from Alibaba Cloud with 35 billion total parameters and 3 billion active parameters per token. It uses a hybrid sparse mixture-of-experts architecture combining Gated...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 235K$0.10/1M $0.90/1M 输出 cache $0.02/1M长文档处理首选 -
#302
Qwen3.6 27B is a dense 27-billion-parameter language model from the Qwen Team at Alibaba, released in April 2026. It features hybrid multimodal capabilities — accepting text, image, and video inputs...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 81K$0.32/1M $3.20/1M 输出长文档处理首选 -
#303
Hy3 preview is a high-efficiency Mixture-of-Experts model from Tencent designed for agentic workflows and production use. It supports configurable reasoning levels across disabled, low, and high modes, allowing it to...
工具 推理上下文 262K 输出 ≤ 235K$0.18/1M $0.60/1M 输出 cache $0.06/1M长文档处理首选 -
#304
MiMo-V2.5-Pro is Xiaomi’s flagship model, delivering strong performance in general agentic capabilities, complex software engineering, and long-horizon tasks, with top rankings on benchmarks such as ClawEval, GDPVal, and SWE-bench Pro....
工具 推理 结构化上下文 262K 输出 ≤ 235K$0.48/1M $1.80/1M 输出 cache $0.10/1M长文档处理首选 -
#305
Kimi K2.6 is Moonshot AI's next-generation multimodal model, designed for long-horizon coding, coding-driven UI/UX generation, and multi-agent orchestration. It handles complex end-to-end coding tasks across Python, Rust, and Go, and...
工具 视觉 推理 结构化上下文 262K 输出 ≤ 235K$0.65/1M $3.41/1M 输出 cache $0.15/1M长文档处理首选 -
#306
Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference — delivering near-31B quality at...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 235K$0.08/1M $0.26/1M 输出 cache $0.04/1M长文档处理首选 -
#307
Gemma 4 26B A4B IT is an instruction-tuned Mixture-of-Experts (MoE) model from Google DeepMind. Despite 25.2B total parameters, only 3.8B activate per token during inference — delivering near-31B quality at...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 32K免费免费可调 · 限时白嫖 -
#308
NVIDIA Nemotron 3 Super is a 120B-parameter open hybrid MoE model, activating just 12B parameters for maximum compute efficiency and accuracy in complex multi-agent applications. Built on a hybrid Mamba-Transformer...
工具 推理 结构化上下文 262K 输出 ≤ 16K$0.07/1M $0.40/1M 输出长文档处理首选 -
#309
Seed-2.0-Lite is a versatile, cost‑efficient enterprise workhorse that delivers strong multimodal and agent capabilities while offering noticeably lower latency, making it a practical default choice for most production workloads across...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 131K$0.25/1M $2.00/1M 输出长文档处理首选 -
#310
Qwen3.5-9B is a multimodal foundation model from the Qwen3.5 family, designed to deliver strong reasoning, coding, and visual understanding in an efficient 9B-parameter architecture. It uses a unified vision-language design...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 40K$0.17/1M $0.25/1M 输出 cache $0.15/1M长文档处理首选 -
#311
Seed-2.0-mini targets latency-sensitive, high-concurrency, and cost-sensitive scenarios, emphasizing fast response and flexible inference deployment. It delivers performance comparable to ByteDance-Seed-1.6, supports 256k context, four reasoning effort modes (minimal/low/medium/high), multimodal understanding,...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 131K$0.10/1M $0.40/1M 输出长文档处理首选 -
#312
The Qwen3.5 Series 35B-A3B is a native vision-language model designed with a hybrid architecture that integrates linear attention mechanisms and a sparse mixture-of-experts model, achieving higher inference efficiency. Its overall...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 81K$0.14/1M $1.00/1M 输出 cache $0.05/1M长文档处理首选 -
#313
The Qwen3.5 27B native vision-language Dense model incorporates a linear attention mechanism, delivering fast response times while balancing inference speed and performance. Its overall capabilities are comparable to those of...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 65K$0.30/1M $2.40/1M 输出长文档处理首选 -
#314
The Qwen3.5 122B-A10B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. In terms of...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 65K$0.26/1M $2.08/1M 输出长文档处理首选 -
#315
The Qwen3.5 series 397B-A17B native vision-language model is built on a hybrid architecture that integrates a linear attention mechanism with a sparse mixture-of-experts model, achieving higher inference efficiency. It delivers...
工具 视觉 视频 推理 结构化上下文 262K 输出 ≤ 81K$0.45/1M $3.00/1M 输出 cache $0.22/1M长文档处理首选 -
#316
Qwen3-Coder-Next is an open-weight causal language model optimized for coding agents and local development workflows. It uses a sparse MoE design with 80B total parameters and only 3B activated per...
工具 结构化上下文 262K 输出 ≤ 235K$0.12/1M $0.80/1M 输出 cache $0.07/1M长文档处理首选 -
#317
Kimi K2.5 is Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm. Built on Kimi K2 with continued pretraining over approximately 15T mixed...
工具 视觉 推理 结构化上下文 262K 输出 ≤ 235K$0.50/1M $2.50/1M 输出 cache $0.15/1M长文档处理首选 -
#318
NVIDIA Nemotron 3 Nano 30B A3B is a small language MoE model with highest compute efficiency and accuracy for developers to build specialized agentic AI systems. The model is fully...
工具 推理 结构化上下文 262K 输出 ≤ 235K$0.06/1M $0.24/1M 输出 cache $0.01/1M长文档处理首选 -
#319
Devstral 2 is a state-of-the-art open-source model by Mistral AI specializing in agentic coding. It is a 123B-parameter dense transformer model supporting a 256K context window. Devstral 2 supports exploring...
工具 文档 结构化上下文 262K 输出 ≤ 209K$0.40/1M $2.00/1M 输出 cache $0.04/1M长文档处理首选 -
#320
The largest model in the Ministral 3 family, Ministral 3 14B offers frontier capabilities and performance comparable to its larger Mistral Small 3.2 24B counterpart. A powerful and efficient language...
工具 视觉 结构化上下文 262K 输出 ≤ 209K$0.20/1M $0.20/1M 输出 cache $0.02/1M长文档处理首选