基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
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。
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
It needs the `prism_ternary` vLLM plugin from fraserprice/bonsai-vllm, which has the run command, kernels and throughput numbers: Built and tested for the RTX PRO 6000 Blackwell only; other NVIDIA GPUs are untested. Problems: open an issue. Produced by `prism-ternary-convert` from the repo above. For benchmarks, sampling parameters and everything else about the model, see the original model card.
译文排队中 · 当前显示来源原文
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
**v3 (`main`, 2026-09-03):** mean full-vocab KLD vs BF16 **0.00556** — lower than the official `Qwen/Qwen3.8-27B-FP8` on the same harness (0.00584) — and the stock **DFlash2 drafter works unchanged** (acceptance length 4.34, identical to a W8A16 target). No residual-stream rotation. 288 of 400 Linear modules run INT8×INT8 on Ampere tensor cores; prefill is 1.42× a W8A16 checkpoint of the same model. Previous revisions stay available: `v2-rotated` (KLD 0.01098, rotation-based; **breaks DFlash2**, MTP only) and `v1-smoothquant-rtn` (0.01414, DFlash2-compatible). If you are here because DFlash2 gave 0% acceptance on v2: `main` fixes it. Teacher-forced full-vocab KLD, BF16 reference captured offline (transformers), candidates served through vLLM (so INT8 *activation* rounding is included). 240 × 512-token WikiText-2 windows = 122,640 positions, top-K 512 with reported tail bound. "Floor-free" is the paired estimator that removes the truncation floor (see `eval/` in the repo).
译文排队中 · 当前显示来源原文
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
The queued trunk quants are uploaded. Each one passed a short Vulkan generation test. `CYBER-FROST-3.8-Q2_K_S.gguf` now carries the MTP head inside the file. The separate `mtp-` files still do not load. Source checkpoint: Blackfrost-AI/CYBER-FROST-3.8-BF16. Architecture `qwen4exp` (`Qwen4ExpForConditionalGeneration`), from `Qwen/Qwen3.8-Flash-Next`. The vision tower is in `mmproj-CYBER-FROST-3.8-f16.gguf`. One mmproj works with every trunk. About 180B parameters, about 6B active. 48 layers, 512 experts, 10 experts per token, one shared expert. Configured context is 262144. Smoke tests used 512 tokens of context. Long context is not tested.
译文排队中 · 当前显示来源原文
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
GGUF conversion of Blackfrost-Research/MiniMax-3.1-Flash-DeRisked, converted with llama.cpp `convert_hf_to_gguf.py`. Vision tower and multimodal projector are included as a separate mmproj GGUF — image understanding works through `llama-mtmd-cli` / `llama-server`. The upstream DeRisked checkpoint ships experts pre-quantized to NVFP4 (compressed-tensors `nvfp4-pack-quantized`) and the rest as MXFP8. This conversion repacks the NVFP4 expert tensors losslessly into GGUF's native NVFP4 layout (per-expert `weight` + `.scale` + `.input_scale` companions) and dequantizes MXFP8 blocks to BF16 in the F16 master before quantizing. *Approximate; experts dominate size at every quant level. All language-model files are split GGUF — download **every** `-0000N-of-0000M` shard of the quant you want.
译文排队中 · 当前显示来源原文
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
> GGUF quants of Blackfrost-AI/RED-SNOW-5.3-FLASH-BF16 - a red-team operations model built on a GLM-5.3-Flash (`glm5_next`) foundation with targeted security adaptation. Full-functionality GGUF exports for llama.cpp: text + **vision included** (the GLM-5.3 vision tower ships as a separate `mmproj` file, keeping every quant pairable with it). Every file was smoke-tested end to end before upload - text runs with `llama-cli`, vision with `llama-server --mmproj`. RED-SNOW 5.3 FLASH is built for authorized offensive-security work: attack-path reasoning, cross-domain operations (identity, cloud, network, app, endpoint, OT), exploit and tradecraft analysis, detection-aware execution, and purple-team handoff. See the base model card for the full picture.
译文排队中 · 当前显示来源原文
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
💜 Qwen Chat    |   🤗 Hugging Face    |   🤖 ModelScope    |    📑 Tech Report    |    📑 Blog    🖥️ Demo    |   💬 WeChat (微信)    |   🫨 Discord   |    Github    We are excited to introduce Qwen-Image-2512, the December update of Qwen-Image’s text-to-image foundational model. You are welcome to try the latest model at Qwen Chat. Compared to the base Qwen-Image model released in August, Qwen-Image-2512 features the following key improvements:
译文排队中 · 当前显示来源原文
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
基础档案已收录,详细介绍后续补充。可先查看厂商、分类与来源。
基础资料已收录 · 缺失资料后续补充
Fast Anime PromptGen generates descriptive safebooru and danbooru tags for anime text-to-image models. This model was trained on 2,470,000 descriptive stable diffusion prompts on the FredZhang7/distilgpt2-stable-diffusion checkpoint for another 4,270,000 steps. Compared to other prompt generation models using GPT2, this one runs with 50% faster forwardpropagation and 40% less disk space & RAM.
译文排队中 · 当前显示来源原文
可以调整厂商、分类或输入输出条件;新模型正持续收录。
调整月度 Token 用量,直观比较不同模型的预估 API 开支。
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。
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。
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。
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。