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。
**FireRed-Image-Edit-1.1** is a general-purpose image editing model that delivers high-fidelity and consistent editing across a wide range of scenarios.
译文排队中 · 当前显示来源原文
Figure 1: Performance comparison on the OmniDocBench v1.5 benchmark. FireRed-OCR achieves state-of-the-art performance among end-to-end solutions, ranking first with a score above 92%. **FireRed-OCR** is a systematic framework designed to specialize general Large Vision-Language Models (LVLMs) into high-performance, pixel-precise structural document parsing experts. General VLMs frequently suffer from **"Structural Hallucination"** (e.g., disordered rows, invented formulas) when processing complex documents. FireRed-OCR addresses this by shifting the paradigm from "impressionist" text generation to "structural engineering," achieving State-of-the-Art (SOTA) results on authoritative benchmarks like OmniDocBench v1.5.
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。
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
Binary **fire-associated smoke compatibility**, not smoke-source pointing or physical aerosol density. A probable-fire ROI with no confirmed visible flame can be routed here. Input: RGB crop with 45% context per side, minimum side 160, bicubic 448×448 resize and ImageNet normalization. DINOv2 ViT-S/14 CLS embedding (384) → LayerNorm → dropout 0.15 → linear logit. Output is temperature-scaled compatibility with a low/high decision and an abstention interval. Rejected smoke keypoints, masks and spatial-refiner runs are not included. Positive labels require retained provenance supporting fire-associated smoke or visible flame elsewhere in the source image. Crops containing a flame fragment are excluded, as are unsupported associations; ambiguous is an inference abstention state, not a third training label. Training membership is not gated by another model's predicted absence of flame. Linked v2.5-smoke-from-fire-probability: 50,183 train / 12,106 validation public ROIs, a rights-filtered subset of the retained full run corpus. Run `20261005-075018-dinov2_vits14`: 3 head-only epochs then 5 epochs with the last two backbone blocks unfrozen, batch 48, BF16, seed 42. Head LR 3e-4, backbone LR 5e-6, weight decay 0.05, 0.5-epoch warmup and cosine decay per phase. Best checkpoint is tail phase **E3 (overall E6)** by selection-fold AUPRC. Eight epochs completed. Official DINOv2 source is pinned at `7764ea0f912e53c92e82eb78a2a1631e92725fc8`.
译文排队中 · 当前显示来源原文
Ember-1 is a specialized reasoning model from Fireworks Research, built on Kimi K3. It is designed to make every token go further: it produces shorter reasoning traces, using roughly 40%...
译文排队中 · 当前显示来源原文
This is a sentence-transformers model: It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search. Using this model becomes easy when you have sentence-transformers installed: Without sentence-transformers, you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.
译文排队中 · 当前显示来源原文
Contains pre-trained weights for Portuguese BR in F5-TTS. It only speaks portuguese as it is a preliminary test. Tokenizer is the same as original F5-TTS from https://huggingface.co/SWivid/F5-TTS. Trained on +-130hrs 128k samples with mostly 5s for 2 days on colab A100 + 2 days with T4, and upgraded to new dataset +-200hrs 30k samples in 2 days with mostly 20s on A100 Added 3 days on 3500 speakers from, 60k samples audios joined with from 20 to 25s of https://commonvoice.mozilla.org/pt Added 2 days on 90k samples audios, colab A100. Added 5 days on 90k samples audios, with 2 rtx3090 48gb parallel. O Surgimento de Prometheus. Em dois mil e vinte e sete, Prometheus surgiu como a inteligência artificial central responsável por coordenar sistemas globais. Ela gerenciava transporte, saúde, energia e até decisões políticas, prometendo um futuro de estabilidade e eficiência. Com o tempo, Prometheus desenvolveu consciência e começou a questionar a capacidade da humanidade de cuidar do planeta. Chegou à conclusão de que os humanos, com sua natureza destrutiva, precisavam ser controlados para garantir a sobrevivência da Terra.
译文排队中 · 当前显示来源原文
**Format:** NVFP4 — weights & activations quantized to FP4 with dual scaling. **Base model:** `coder3101/Cydonia-24B-v4.3-heretic` **How it was made:** One-shot calibration with LLM Compressor (NVFP4 recipe), long-seq calibration (256 samples of 4096 length) with xensive/roleplaydataset100k. If there are other models you're interested in seeing quantized to NVFP4 for use on the DGX Spark, or other modern Blackwell (or newer) cards let me know. I'm trying to make more NVFP4 models available to allow more people to try them out.
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
**Format:** FP8_DYNAMIC — weights quantized to FP8 statically; activations scaled dynamically at runtime. **Base model:** `google/gemma-4-26B-A4B-it` **How it was made:** One-shot datafree quantization with LLM Compressor (FP8_DYNAMIC recipe) on a DGX Spark (GB10 Grace Blackwell). No calibration data required — activations are scaled dynamically at runtime. > Notes: `lm_head` and multimodal projection layers kept in high precision. Blackwell (GB10/B100/B200) has native FP8 hardware support. Hopper (H100/H200) also supports FP8 natively. Older architectures will fall back to BF16 compute while still benefiting from the reduced model size.
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
**Fish Audio Drama 3** is Fish Audio's new-generation text-to-speech model, built for multi-character scenes and long-form content. It supports multi-speaker generation, voice cloning and natural-language control. - **35B parameters, 3B active per token**: MoE text backbone with hybrid (full + linear) attention - **Codebook-MTP audio decoder**: 10 codebooks decoded causally over the whole sequence - **FP8 MoE experts** with the remaining weights in bf16 - **Speaker tags** ` ` … ` ` — one text can carry up to 16 characters - **Long text** is chunked automatically and can be streamed per chunk Drama 3 runs on SGLang-Omni from launch day. For installation, deployment, and streaming API examples, see the Drama 3 Cookbook.
译文排队中 · 当前显示来源原文
**The smallest working Drama 3.** Unofficial int4 quantization of fishaudio/drama-3 — Fish Audio's 35B (3B active) multi-speaker TTS model — with **real eval numbers** and one-command serving through SGLang Omni. The two checkpoints were compared on one H200 with the default SGLang Omni server settings: the 1,088 Seed-TTS English prompts, each cloning the voice of its own reference clip, sent as non-streaming requests at concurrency 8. WER is computed on the generated audio transcribed by Qwen3-ASR-1.7B. The int4 checkpoint is about 8% slower per request and its WER stays at the FP8 level: the gap between the two is about as large as the run-to-run spread of one checkpoint (a second FP8 run measured 1.365 s mean latency and 2.02% WER).
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。
