跳到主内容
智客 ZICQ

逻辑推理榜

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:25 分钟前(2026-11-18 10:41)
清除
数据来源与评测方法

声明支持 reasoning 推理 effort 的模型,按 Intelligence 指数排序。

📊 评测基准与方法论说明
1. 数据来源:模型元数据与调用价格每小时自 OpenRouter (openrouter.ai) 官方接口同步,包含商业模型、开源权重及实验预览版本。
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
当前榜单模型数 478
免费可调 17
支持视觉 334
支持工具调用 415
支持推理 361
最长上下文 xAI: Grok 4.20 Multi-Agent 2M tokens
已知输入价格均值 $1.55 / 1M 美元 / 百万 token
最低价模型 Z.ai: GLM 5.2 $0.007 / 1M
当前维度 TOP · 智能指数 Anthropic: Claude Fable 5.2 64.6
⚡ 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 Anthropic: Claude Fable 5.2 Anthropic Anthropic: Claude Fable 5.2 (batch) Anthropic OpenAI: GPT-6.2 Astra OpenAI
综合智能 / 编码 / 智能体 64.6 / 0.0 / 0.0 64.6 / 0.0 / 0.0 62.1 / 0.0 / 0.0
上下文窗口 1M tokens 1M tokens 1.1M tokens
输入 / 输出单价 (1M) $8.00 / $40.00 $4.00 / $20.00 $8.00 / $40.00
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
  1. #81
    Thinking Machines: Inkling Small (batch) Thinkingmachines 视听多模态 有基准

    Inkling Small is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 12B active parameters out of 276B total. It is positioned as the smaller, more efficient member of...

    工具 视觉 音频 推理
    智能 12.9
    上下文 524K 输出 ≤ 262K
    $0.36/1M $0.96/1M 输出 cache $0.08/1M
    推理能力稳健
  2. #82
    Thinking Machines: Inkling Thinkingmachines 视听多模态 有基准

    Inkling is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 41B active parameters out of 975B total. It is designed for general-purpose reasoning, coding, agentic and tool-use systems,...

    工具 视觉 音频 推理
    智能 12.5
    上下文 524K 输出 ≤ 471K
    $1.00/1M $4.05/1M 输出 cache $0.17/1M
    推理能力稳健
  3. #83
    NVIDIA: Nemotron 3 Ultra NVIDIA 通用大语言模型 有基准

    NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...

    工具 推理 结构化
    智能 11.8
    上下文 262K 输出 ≤ 16K
    $0.50/1M $2.20/1M 输出 cache $0.10/1M
    推理能力稳健
  4. #84
    NVIDIA: Nemotron 3 Ultra (free) NVIDIA 通用大语言模型 免费 有基准

    NVIDIA Nemotron 3 Ultra is an open frontier-reasoning and orchestration model from NVIDIA, with 55B active parameters out of 550B total (MoE). Built on a hybrid Transformer-Mamba mixture-of-experts architecture, it...

    工具 推理
    智能 11.8
    上下文 1M 输出 ≤ 65K
    免费
    免费可调 · 限时白嫖
  5. #85
    Google: Gemini 3.5 Flash Lite Google DeepMind 音视频多模态 有基准

    Gemini 3.5 Flash Lite is a high-efficiency model from Google with upgraded agentic capabilities. It is suited for subagents that execute focused tasks within complex, multi-agent workflows.

    工具 视觉 音频 视频 文档 推理 结构化
    智能 11.7
    上下文 1.0M 输出 ≤ 65K
    $0.30/1M $2.50/1M 输出 cache $0.03/1M
    推理能力稳健
  6. #86
    Google: Gemini 3.5 Flash Lite (batch) Google DeepMind 音视频多模态 有基准

    Gemini 3.5 Flash Lite is a high-efficiency model from Google with upgraded agentic capabilities. It is suited for subagents that execute focused tasks within complex, multi-agent workflows.

    工具 视觉 音频 视频 文档 推理 结构化
    智能 11.7
    上下文 1.0M 输出 ≤ 65K
    $0.15/1M $1.25/1M 输出 cache $0.01/1M
    推理能力稳健
  7. #87
    Anthropic: Claude Haiku 4.5 Anthropic 文档视觉模型 有基准

    Claude Haiku 4.5 is Anthropic’s fastest and most efficient model, delivering near-frontier intelligence at a fraction of the cost and latency of larger Claude models. Matching Claude Sonnet 4’s performance...

    工具 视觉 文档 推理 结构化
    智能 7.4
    上下文 200K 输出 ≤ 64K
    $1.00/1M $5.00/1M 输出 cache $0.10/1M
    推理能力稳健
  8. #88
    Anthropic: Claude Haiku 4.5 (batch) Anthropic 文档视觉模型 有基准

    Claude Haiku 4.5 is Anthropic’s fastest and most efficient model, delivering near-frontier intelligence at a fraction of the cost and latency of larger Claude models. Matching Claude Sonnet 4’s performance...

    工具 视觉 文档 推理 结构化
    智能 7.4
    上下文 200K 输出 ≤ 64K
    $0.50/1M $2.50/1M 输出 cache $0.05/1M
    推理能力稳健
  9. #89
    Meta: Muse Glimmer 30B Meta 视觉多模态 有基准

    Muse Glimmer 30B is a dense, open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark and optimized for autonomous agents on consumer hardware. It is suited for long-horizon...

    工具 视觉 推理 结构化
    智能 6.3
    上下文 131K 输出 ≤ 117K
    $0.35/1M $1.50/1M 输出 cache $0.04/1M
    推理能力稳健
  10. #90
    Meta: Muse Glimmer 30B (batch) Meta 视觉多模态 有基准

    Muse Glimmer 30B is a dense, open-weight multimodal model from Meta Superintelligence Labs, distilled from Muse Spark and optimized for autonomous agents on consumer hardware. It is suited for long-horizon...

    工具 视觉 推理 结构化
    智能 6.3
    上下文 131K 输出 ≤ 16K
    $0.24/1M $0.96/1M 输出 cache $0.03/1M
    推理能力稳健
  11. #91
    Cohere: Command A+ Cohere 视觉多模态 有基准

    Command A+ is Cohere's flagship model for enterprise agentic workflows. It accepts text and image inputs with a 192K context window, supports native tool calling with strict tool schemas, structured...

    工具 视觉 推理 结构化
    智能 6.2
    上下文 192K 输出 ≤ 64K
    $0.30/1M $1.50/1M 输出 cache $0.15/1M
    推理能力稳健
  12. #92
    Google: Gemini 3.1 Flash Lite Preview Google DeepMind 音视频多模态 有基准

    Gemini 3.1 Flash Lite Preview is Google's high-efficiency model optimized for high-volume use cases. It outperforms Gemini 2.5 Flash Lite on overall quality and approaches Gemini 2.5 Flash performance across...

    工具 视觉 音频 视频 文档 推理 结构化
    智能 6.2
    上下文 1.0M 输出 ≤ 65K
    $0.25/1M $1.50/1M 输出 cache $0.02/1M
    推理能力稳健
  13. #93
    LiquidAI: LFM3-1.7B Liquid 通用大语言模型 有基准

    A compact 1.7B-parameter text model with a 256K-token context, post-trained for agents, tool calling, and structured outputs. It reasons before it answers. Prompts and outputs may be retained and used...

    工具 推理 结构化
    智能 6.0
    上下文 262K 输出 ≤ 8K
    $0.08/1M $0.20/1M 输出 cache $0.02/1M
    推理能力稳健
  14. #94
    LiquidAI: LFM3-1.7B (free) Liquid 通用大语言模型 免费 有基准

    A compact 1.7B-parameter text model with a 256K-token context, post-trained for agents, tool calling, and structured outputs. It reasons before it answers. Prompts and outputs may be retained and used...

    工具 推理 结构化
    智能 6.0
    上下文 262K 输出 ≤ 8K
    免费
    免费可调 · 限时白嫖
  15. #95
    NVIDIA: Nemotron 3.5 Lightning NVIDIA 通用大语言模型 有基准

    NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model from NVIDIA, with 3B active parameters out of 30B total. It is suited for high-throughput agentic workloads and specialized tasks that...

    工具 推理 结构化
    智能 5.5
    上下文 1.0M 输出 ≤ 235K
    $0.07/1M $0.20/1M 输出 cache $0.04/1M
    推理能力稳健
  16. #96
    NVIDIA: Nemotron 3.5 Lightning (free) NVIDIA 通用大语言模型 免费 有基准

    NVIDIA Nemotron 3.5 Lightning is an open mixture-of-experts model from NVIDIA, with 3B active parameters out of 30B total. It is suited for high-throughput agentic workloads and specialized tasks that...

    工具 推理
    智能 5.5
    上下文 1M 输出 ≤ 65K
    免费
    免费可调 · 限时白嫖
  17. #97
    IBM: Granite 4.2 8B Ibm-Granite 通用大语言模型 有基准

    Granite 4.2 8B is a dense reasoning model from IBM. It is suited for mathematics, code generation, multilingual dialogue, and agentic workflows that need multi-step reasoning. It supports full, low-effort,...

    工具 推理 结构化
    智能 5.1
    上下文 131K 输出 ≤ 117K
    $0.06/1M $0.15/1M 输出 cache $0.01/1M
    推理能力稳健
  18. #98
    Mistral: Mistral Medium 3.5 Mistral AI 文档视觉模型 有基准

    Mistral Medium 3.5 is a dense 128B instruction-following model from Mistral AI. It supports text and image inputs with text output, and is designed for agentic workflows, coding, and complex...

    工具 视觉 文档 推理 结构化
    智能 4.6
    上下文 262K 输出 ≤ 209K
    $1.50/1M $7.50/1M 输出
    推理能力稳健
  19. #99
    Mistral: Mistral Medium 3.5 (batch) Mistral AI 文档视觉模型 有基准

    Mistral Medium 3.5 is a dense 128B instruction-following model from Mistral AI. It supports text and image inputs with text output, and is designed for agentic workflows, coding, and complex...

    工具 视觉 文档 推理 结构化
    智能 4.6
    上下文 262K 输出 ≤ 209K
    $0.75/1M $3.75/1M 输出
    推理能力稳健
  20. #100
    Google: Gemma 4 31B Google DeepMind 视觉多模态 有基准

    Gemma 4 31B Instruct is Google DeepMind's 30.7B dense multimodal model supporting text and image input with text output. Features a 256K token context window, configurable thinking/reasoning mode, native function...

    工具 视觉 视频 推理 结构化
    智能 4.4
    上下文 262K 输出 ≤ 16K
    $0.20/1M $0.40/1M 输出
    推理能力稳健