视觉多模态榜
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:5 分钟前(2026-11-21 06:34)数据来源与评测方法
支持图像 / 视频 / 文档输入的多模态模型,排名由 Intelligence 指数 + 视觉能力加成。
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 | 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 |
-
#401
The relace-search model uses 4-12 `view_file` and `grep` tools in parallel to explore a codebase and return relevant files to the user request. In contrast to RAG, relace-search performs agentic...
工具 结构化上下文 256K 输出 ≤ 128K$1.00/1M $3.00/1M 输出图像理解稳定 -
#402
DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...
工具 推理 结构化上下文 163K 输出 ≤ 147K$0.26/1M $0.38/1M 输出 cache $0.13/1M图像理解稳定 -
#403
Kimi K2 Thinking is Moonshot AI’s most advanced open reasoning model to date, extending the K2 series into agentic, long-horizon reasoning. Built on the trillion-parameter Mixture-of-Experts (MoE) architecture introduced in...
工具 推理 结构化上下文 262K 输出 ≤ 98K$0.60/1M $2.50/1M 输出 cache $0.15/1M图像理解稳定 -
#404
Voxtral Small is an enhancement of Mistral Small 3, incorporating state-of-the-art audio input capabilities while retaining best-in-class text performance. It excels at speech transcription, translation and audio understanding. Input audio...
工具 音频 文档 结构化上下文 32K 输出 ≤ 26K$0.10/1M $0.30/1M 输出 cache $0.01/1M图像理解稳定 -
#405
gpt-oss-safeguard-20b is a safety reasoning model from OpenAI built upon gpt-oss-20b. This open-weight, 21B-parameter Mixture-of-Experts (MoE) model offers lower latency for safety tasks like content classification, LLM filtering, and trust...
工具 推理 结构化上下文 131K 输出 ≤ 65K$0.07/1M $0.30/1M 输出 cache $0.04/1M图像理解稳定 -
#406
Compared with GLM-4.5, this generation brings several key improvements: Longer context window: The context window has been expanded from 128K to 200K tokens, enabling the model to handle more complex...
工具 推理 结构化上下文 204K 输出 ≤ 131K$0.50/1M $2.00/1M 输出 cache $0.10/1M图像理解稳定 -
#407
DeepSeek-V3.2-Exp is an experimental large language model released by DeepSeek as an intermediate step between V3.1 and future architectures. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...
工具 推理 结构化上下文 163K 输出 ≤ 147K$0.27/1M $0.41/1M 输出图像理解稳定 -
#408
Uncensored and creative writing model based on Mistral Small 3.2 24B with good recall, prompt adherence, and intelligence.
结构化上下文 131K 输出 ≤ 117K$0.30/1M $0.50/1M 输出 cache $0.15/1M图像理解稳定 -
#409
Relace Apply 3 is a specialized code-patching LLM that merges AI-suggested edits straight into your source files. It can apply updates from GPT-4o, Claude, and others into your files at...
上下文 256K 输出 ≤ 128K$0.85/1M $1.25/1M 输出图像理解稳定 -
#410
DeepSeek-V3.1 Terminus is an update to [DeepSeek V3.1](/deepseek/deepseek-chat-v3.1) that maintains the model's original capabilities while addressing issues reported by users, including language consistency and agent capabilities, further optimizing the model's...
工具 推理 结构化上下文 163K 输出 ≤ 147K$0.27/1M $1.00/1M 输出图像理解稳定 -
#411
Qwen3 Coder Flash is Alibaba's fast and cost efficient version of their proprietary Qwen3 Coder Plus. It is a powerful coding agent model specializing in autonomous programming via tool calling...
工具 结构化上下文 1M 输出 ≤ 65K$0.20/1M $0.97/1M 输出 cache $0.04/1M图像理解稳定 -
#412
Qwen3-Next-80B-A3B-Instruct is an instruction-tuned chat model in the Qwen3-Next series optimized for fast, stable responses without “thinking” traces. It targets complex tasks across reasoning, code generation, knowledge QA, and multilingual...
工具 结构化上下文 262K 输出 ≤ 235K$0.10/1M $1.10/1M 输出 cache $0.07/1M图像理解稳定 -
#413
Kimi K2 0905 is the September update of [Kimi K2 0711](moonshotai/kimi-k2). It is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32...
工具 结构化上下文 262K 输出 ≤ 98K$0.60/1M $2.50/1M 输出图像理解稳定 -
#414
Hermes 4 is a large-scale reasoning model built on Meta-Llama-3.1-405B and released by Nous Research. It introduces a hybrid reasoning mode, where the model can choose to deliberate internally with...
推理 结构化上下文 131K 输出 ≤ 117K$1.00/1M $3.00/1M 输出 cache $0.20/1M图像理解稳定 -
#415
DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context...
工具 推理 结构化上下文 163K 输出 ≤ 147K$0.25/1M $0.95/1M 输出 cache $0.13/1M图像理解稳定 -
#416
gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for...
工具 推理 结构化上下文 131K 输出 ≤ 117K$0.02/1M $0.08/1M 输出 cache $0.01/1M图像理解稳定 -
#417
gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for...
工具 推理 结构化上下文 131K 输出 ≤ 117K$0.02/1M $0.11/1M 输出图像理解稳定 -
#418
Mistral's cutting-edge language model for coding released end of July 2025. Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. [Blog Post](https://mistral.ai/news/codestral-25-08)
工具 文档 结构化上下文 256K 输出 ≤ 204K$0.30/1M $0.90/1M 输出 cache $0.03/1M图像理解稳定 -
#419
Qwen3-Coder-30B-A3B-Instruct is a 30.5B parameter Mixture-of-Experts (MoE) model with 128 experts (8 active per forward pass), designed for advanced code generation, repository-scale understanding, and agentic tool use. Built on the...
工具 结构化上下文 262K 输出 ≤ 235K$0.15/1M $0.60/1M 输出图像理解稳定 -
#420
Qwen3-30B-A3B-Instruct-2507 is a 30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference. It operates in non-thinking mode and is designed for high-quality instruction following, multilingual understanding, and...
工具 结构化上下文 262K 输出 ≤ 235K$0.10/1M $0.30/1M 输出 cache $0.10/1M图像理解稳定