AI NEWS.ccAI正在发生 未来正在更新
♡ 我的收藏 0
AI NEWS · 模型情报站发现能力边界，让每一次选择更有依据 ↗基础档案先收录 · 缺失资料持续补齐
THE AI MODEL DIRECTORY 模型探索，从这里开始

找到适合你的 AI 模型。

从语言到影像，从开源到 API。读懂模型能力、比较真实成本，
为你的下一个想法，找到合适的智能引擎。

热门探索
84,991 款模型档案13202 家厂商 / 发布者45 个能力分类465 款有可比报价 · 30 款有公开评测
T文本生成▧图像生成▷视频生成⌘代码生成◖语音合成♫音乐生成⬡3D 建模▤文档分析
DISCOVER & COMPARE

探索 AI 模型 22

为你发现值得关注的模型

The full NVFP4 build of the LTX-2.5 22B distilled transformer, stamped so ComfyUI actually loads it. Smallest working 22B on the board — **18.7GB, ~2.8GB lighter than the official comfy-int8** — and the same-seed output holds up frame for frame. Top row: text-to-video. Bottom row: **image-to-video** from a single keyframe — same start image, same seed, both quants. Native synced audio on all four (yes, the disco clips come with their own disco track). Lightricks ships an NVFP4 export of the 2.5 distilled DiT, but it's a raw TensorRT-ModelOpt export. Load it through ComfyUI's UNETLoader and you get:

译文排队中 · 当前显示来源原文

视频生成
输入 文本→ 输出 视频
价格待核验多模态与渠道费用须分别核对
查看详情 ↗
来源 / 文档 ↗

**This is an independent NVFP4 quantization of Qwen Image 2.1. The Qwen team created the original model.** The original model and architecture are Qwen's work. I converted the official Comfy-Org/Qwen-Image-2.1 BF16 checkpoint package and packaged the quantized weights, workflows, and validation here. This is not a new base model or an official Qwen release. I converted Qwen Image 2.1's image transformer and Qwen3-VL 8B language encoder directly from the official BF16 weights into native ComfyUI NVFP4. The vision tower, embeddings, output head, critical image-model tensors, and VAE stay BF16. Real Blackwell FP4 kernels in both denoising and text conditioning, with an included small runtime patch for the latter. No new nodes. **RTX 5090: 6.746 s median for a complete 1024×1024, 40-step run** versus 15.629 s with BF16 and 7.553 s with official INT8. Three warm runs per variant, every node executed, including text encoding. This one-prompt benchmark used the encoder patch; full method and limits.

译文排队中 · 当前显示来源原文

图像生成
输入 文本→ 输出 图像
价格待核验多模态与渠道费用须分别核对
查看详情 ↗
来源 / 文档 ↗

Native ComfyUI weights, a selectively quantized image transformer and encoder, and seven ready-to-load workflows. Fashion, casual photography, editing, multiple references, typography, illustration, and transparent PNGs. **No custom nodes or core patches required.** All seven workflows were tested on unmodified ComfyUI 0.39.0 with custom nodes disabled. Update ComfyUI and its normal dependencies, then place these files in the matching folders: **15.70 GB total — 51.6% smaller than the original BF16 package.** These are file sizes. Model checksums are included.

译文排队中 · 当前显示来源原文

图像生成
输入 文本→ 输出 图像
价格待核验多模态与渠道费用须分别核对
查看详情 ↗
来源 / 文档 ↗

V4 still rocks as the original Qwen3 prompt-engineer build. The new V6 release is the native Z-Image Turbo text-encoder build, rebuilt with SMART DoRA and trained locally for ComfyUI drop-in use plus LM Studio prompt enhancement. This is Z-Engineer V4, the culmination of extensive research into what makes an AI prompt engineer actually *good* at its job. Built on the Qwen 3 architecture and trained using a novel **SMART Training** methodology, this 4B parameter model doesn't just describe scenes—it understands the craft of visual storytelling down to the lens flare. Z-Engineer V4 is a fully fine-tuned (not LoRA, we went *all in*) version of the text encoder from Tongyi-MAI/Z-Image-Turbo. It's been specifically trained to understand the nuances of AI Image Generation workflows.

译文排队中 · 当前显示来源原文

文本生成
输入 文本→ 输出 文本
价格待核验多模态与渠道费用须分别核对
查看详情 ↗
来源 / 文档 ↗

This is the **abliterated** version of the Z-Image-Turbo text encoder. I took the **p-e-w heretic method** and hammered it through **1000 trials**, specifically targeting **BOTH** image generation refusals and those peskier general refusals. The result? - **KL Divergence:** A tiny `0.0004` (basically no lobotomy! 🧠✨) - **Refusal Rate:** Only `4/100` in my torture tests. Brought to you by the same chaotic good energy behind: - Z-Image-Engineer-V2.5 - The original Z-Image-Engineer

译文排队中 · 当前显示来源原文

分类待补充
输入 待核验→ 输出 待核验
价格待核验多模态与渠道费用须分别核对
查看详情 ↗
来源 / 文档 ↗

The main repo contains the merged HF safetensors. This repo contains the quant ladder for the ComfyUI-Z-Engineer node, LM Studio, ComfyUI `CLIPLoaderGGUF`, llama.cpp-style loaders, and local prompt-enhancement workflows. **Z-Image-Engineer V6** is a SMART DoRA fine-tuned 4B Qwen text encoder from `Tongyi-MAI/Z-Image-Turbo`. The comparison examples were generated from direct LM Studio user requests like this, with no separate system prompt. `V6_SYSTEM_PROMPT.md` is included only as an optional preset for people who want a stricter prompt-only chat setup.

译文排队中 · 当前显示来源原文

文本生成
输入 文本→ 输出 文本
价格待核验多模态与渠道费用须分别核对
查看详情 ↗
来源 / 文档 ↗
COST EXPLORER

好模型，也要算好成本。

费用分析

调整月度 Token 用量，直观比较不同模型的预估 API 开支。

1 M = 100 万 Token · 美元计价
仅比较同一渠道的基础文本 Token 报价

使用已采集且单位可比的 USD / 百万 Token 渠道报价。报价为采集时快照；不计缓存优惠、阶梯价、推理额外输出与税费，请以来源的当前报价为准。

MODEL HANDBOOK

读懂模型，再做选择

选型与常见问题
API 价格应该怎么比较？+

先区分输入、输出和缓存价格，再使用同一任务和用量估算总费用。图像、音频和视频可能按张、时长或分辨率计费，不能直接与 Token 单价比较。

开放权重，就代表可以免费商用吗？+

不一定。必须逐一核对具体模型版本的许可证、商用限制及使用政策。自己部署还需要考虑显存、推理资源和运维费用。

排行榜第一，就是最适合我的模型吗？+

榜单取决于测试任务、版本与评测方法。建议用实际业务样本比较准确性、时延、稳定性和成本，再选择模型。

AI NEWS · 从信息到选择

资讯看趋势，工具提效率，模型筑可能。

探索 AI 资讯 ↗发现 AI 工具 ↗