追踪 DeepSeek 模型发布、推理能力、开放 API 与官方更新。
test-llama-archs : generate dummy test vocab (#29084) Assisted-by: pi:llama.cpp/DeepSeek-V4-Flash-Vision-Exp **Website:** - **Attestations:** - **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b11049/llama-
### Summary of major features and improvements * #### More GenAI coverage and framework integrations to minimize code changes * New models supported: * On CPU: Gemma-3n * On CPU, GPU: Kokoro-82M, Qwen3-VL-4B with EAGLE-3, Qwen3-ASR, Muse Glimmer 30B, Qwen3.8 2
Mixture of experts (MoE) has become one of the defining architectural trends in large-scale AI model training. DeepSeek, Qwen, and Mixtral are examples of MoE...
## Verify Docker Image Signature All LiteLLM Docker images are signed with [cosign](https://docs.sigstore.dev/cosign/overview/). Every release is signed with the same key introduced in [commit `0112e53`](https://github.com/BerriAI/litellm/commit/0112e53046018d
# v0.29.0 ## Highlights This release features 594 commits from 277 contributors (91 new)! * **Model Runner V2 is now the default for all models** (#53183), completing the rollout that began with pooling models (#48290). MRV2 also gained CUDA graph memory profi
ONNX Runtime 1.30.0 expands generative AI inference, improves CPU and GPU performance, adds Go bindings, and strengthens runtime reliability. These notes cover changes since ONNX Runtime 1.29.1. ## Highlights - Expanded CUDA inference support with variable-len
# Release v5.17.0 ## New Model additions ### HYV4 Hy4-Preview is a 780B-parameter mixture-of-experts language model that activates 49B parameters per token. Each MoE layer holds 256 routed experts plus one always-active shared expert and routes every token to
模型仓库动态。仓库创建：2026-09-10T02:17:58.000Z。最后修改：2026-09-10T08:18:10.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。 任务：image-text-to-text。
来源标注日期：2026-09-10（未提供具体时刻）。 Today, we officially release the DeepSeek-V4.1-Flash model. It is the smallest model in our new architecture family, with native multimodal visual understanding. The new architecture is designed for a higher capability ceiling, fast
模型仓库动态。仓库创建：2026-08-31T06:16:18.000Z。最后修改：2026-09-01T09:22:10.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。 任务：image-text-to-text。
# v0.28.0 ## Highlights This release features 584 commits from 270 contributors (76 new)! * **Kimi-K3 performance push**: a major optimization effort for Kimi-K3 across the stack — Decode Context Parallel (DCP) support (#50484), fused FlashKDA decode and prefi
## What's Changed ### Claude Desktop Developers can now easily configure Claude Desktop to seamlessly work with Ollama as a third-party gateway provider. ### Improved caching * Fixed a hang where agent clients that cancel long prefills * Prefill restore points
# Highlights *710 PRs from 212 contributors.* **New models in this release** (see the [cookbook](https://docs.sglang.io/cookbook) for all supported models): | Model | Type | PRs | Cookbook | |---|---|---|---| | Muse Glimmer | Autoregressive (Multimodal) | [#34
来源标注日期：2026-08-21（未提供具体时刻）。 Today, the new multimodal vision understanding model DeepSeek-V4-Flash-Vision-Exp is now available on the DeepSeek API platform. This is an experimental model that can be accessed by setting model='deepseek-v4-flash-vision-exp'. Ter
模型仓库动态。仓库创建：2026-08-13T03:05:06.000Z。最后修改：2026-08-13T16:28:28.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。 任务：text-generation。
来源标注日期：2026-08-13（未提供具体时刻）。 The GA release of DeepSeek-V4-Pro has been rolled out on the APP, Web, and API. The API calling method remains unchanged — simply set the model name to deepseek-v4-pro to use the latest version. Significantly enhanced Agent capabili
# vLLM v0.27.0 Release Notes ## Highlights This release features 561 commits from 242 contributors (64 new)! * **Kimi K3 support** with a full stack landing in one release: core model files and kernels (#50089, #50000), Python (#50093) and Rust (#50104) fronte
模型仓库动态。仓库创建：2026-07-31T07:30:24.000Z。最后修改：2026-08-01T03:07:41.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。 任务：text-generation。
来源标注日期：2026-07-31（未提供具体时刻）。 The official release of the DeepSeek-V4-Flash API is now in public beta. The API calling method remains unchanged — simply set the model name to deepseek-v4-flash to use the latest version. Significantly enhanced agent capabilities,
# vLLM v0.26.0 Release Notes ## Highlights This release features 411 commits from 212 contributors (61 new)! * **New Inkling model family** with a full support stack: base modeling (#48799), piecewise CUDA graph support (#48822), Hopper FA4 relative attention
# Highlights *574 PRs from 169 contributors.* **DSpark: confidence-driven speculative decoding**: A new speculative algorithm. It drafts semi-autoregressively in blocks, then sizes each verify window from the draft's own confidence instead of a fixed draft len
# vLLM v0.25.0 Release Notes ## Highlights This release features 558 commits from 232 contributors (64 new)! * **Model Runner V2 is now the default for all dense models** (#44443). Building on quantized-model support from the previous release, MRv2 is now the
模型仓库动态。仓库创建：2026-06-27T02:27:36.000Z。最后修改：2026-07-04T03:15:12.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。 任务：text-generation。
模型仓库动态。仓库创建：2026-06-27T03:02:56.000Z。最后修改：2026-07-04T03:14:46.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。 任务：text-generation。
# vLLM v0.24.0 Release Notes ## Highlights This release features 571 commits from 256 contributors (77 new)! * **MiniMax-M3**: Added support for the new **MiniMax-M3** model (#45381), with a fast follow-on of BF16/FP8 indexer via MSA (#45892), MXFP4 support (#
模型仓库动态。仓库创建：2026-06-28T12:37:17.000Z。最后修改：2026-06-28T12:38:12.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。
模型仓库动态。仓库创建：2026-06-28T12:36:20.000Z。最后修改：2026-06-28T12:37:15.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。
模型仓库动态。仓库创建：2026-06-28T12:35:29.000Z。最后修改：2026-06-28T12:36:18.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。
模型仓库动态。仓库创建：2026-06-28T12:34:59.000Z。最后修改：2026-06-28T12:35:27.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。
模型仓库动态。仓库创建：2026-06-28T12:33:24.000Z。最后修改：2026-06-28T12:34:57.000Z。仓库修改不等同于模型正式发布。 数据经第三方 Hugging Face 镜像采集，需以原始模型页面复核。
