AI 项目版本更新

<details open> ggml : add a set of functions for checking contiguity of inner tensor dimensions (#25650) Co-authored-by: Stanisław Szymczyk <sszymczy@gmail.com> </details> **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-macos-arm64.tar.gz) - macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780) - [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-xcframework.zip) **Linux:** - [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-ubuntu-x64.tar.gz) - [Ubuntu arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-ubuntu-arm64.tar.gz) - [Ubuntu s390x (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-ubuntu-s390x.tar.gz) - [Ubuntu x64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-ubuntu-vulkan-x64.tar.gz) - [Ubuntu arm64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-ubuntu-vulkan-arm64.tar.gz) - [Ubuntu x64 (ROCm 7.2)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-ubuntu-rocm-7.2-x64.tar.gz) - [Ubuntu x64 (OpenVINO)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-ubuntu-openvino-2026.2.1-x64.tar.gz) - [Ubuntu x64 (SYCL FP32)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-ubuntu-sycl-fp32-x64.tar.gz) - [Ubuntu x64 (SYCL FP16)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-ubuntu-sycl-fp16-x64.tar.gz) **Android:** - [Android arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-android-arm64.tar.gz) **Windows:** - [Windows x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-win-cpu-x64.zip) - [Windows arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-win-cpu-arm64.zip) - [Windows arm64 (OpenCL Adreno)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-win-opencl-adreno-arm64.zip) - [Windows x64 (CUDA 12)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-win-cuda-12.4-x64.zip) - [CUDA 12.4 DLLs](https://github.com/ggml-org/llama.cpp/releases/download/b10002/cudart-llama-bin-win-cuda-12.4-x64.zip) - [Windows x64 (CUDA 13)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-win-cuda-13.3-x64.zip) - [CUDA 13.3 DLLs](https://github.com/ggml-org/llama.cpp/releases/download/b10002/cudart-llama-bin-win-cuda-13.3-x64.zip) - [Windows x64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b10002/llama-b10002-bin-win

## 1.46.1 (2026-07-14) <!-- Release notes generated using configuration in .github/release.yml at main --> ## What's Changed ### Bug Fixes * fix(app-server): preserve conversation created_at across lifecycle webhooks by @Sujit-1509 in https://github.com/OpenHands/OpenHands/pull/15243 * fix(mcp): preserve SaaS credentials with encrypted storage by @simonrosenberg in https://github.com/OpenHands/OpenHands/pull/15257 ### Maintenance * ci: wait for the docker build before retagging images by @jlav in https://github.com/OpenHands/OpenHands/pull/15213 * chore: bump SDK packages to v1.35.0 by @all-hands-bot in https://github.com/OpenHands/OpenHands/pull/15240 * chore: bump SDK packages to v1.36.0 by @all-hands-bot in https://github.com/OpenHands/OpenHands/pull/15256 ## New Contributors * @Sujit-1509 made their first contribution in https://github.com/OpenHands/OpenHands/pull/15243 **Full Changelog**: https://github.com/OpenHands/OpenHands/compare/cloud-1.46.0...cloud-1.46.1 --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please).

# vLLM v0.25.1 ## Highlights This release features 2 commits from 2 contributors (1 new)! v0.25.1 is a patch release containing two targeted bug fixes on top of v0.25.0. ### Bug Fixes * **Avoid blocking model launching when no system FFmpeg is available for TorchCodec** (#47888). Previously `import torchcodec` raised a `RuntimeError` at import time when system FFmpeg was missing, which blocked startup (e.g. `vllm serve Qwen/Qwen3-VL-2B-Instruct`) even when TorchCodec was not in use. The error is now deferred to runtime so it only surfaces if TorchCodec is actually needed. * **Guard mixed-dtype allreduce RMSNorm quant fusions** (#48330). The fused FlashInfer allreduce + RMSNorm + static-quantization patterns could match graphs where the activation and RMSNorm weight dtypes differ (e.g. a BF16 residual stream with an FP32 Gemma/Qwen-style RMSNorm weight in NVFP4 models), corrupting the hidden state and producing garbage output such as repeated `!!!!!` tokens. A dtype-match guard now routes incompatible mixed-dtype graphs to the safe path, while same-dtype models retain the full allreduce + RMSNorm + quant fusion. ## Contributors @Isotr0py, @hugo-cen ## New Contributors * @hugo-cen made their first contribution in https://github.com/vllm-project/vllm/pull/48330

- Added a manual API key escape hatch for Cline OAuth providers, so you can enter a key by hand from settings - Fixed provider config not reloading when switching models - Fixed auto-update failing to detect Bun global installs after symlink resolution - Fixed unexpected logouts caused by transient network or server errors during token refresh - The ClinePass usage-limit error is now surfaced clearly when you hit the limit - Session id is now preserved when continuing within the same session - Hardened context compaction budget handling **Full Changelog**: https://github.com/cline/cline/compare/cli-v3.0.39...cli-v3.0.40

## Mem0 Node CLI (v0.2.11) **Bug Fixes:** - **Platform backend:** Encode dynamic URL path segments so memory and entity IDs containing special characters no longer produce malformed requests ([#5963](https://github.com/mem0ai/mem0/pull/5963))

## 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/0112e53046018d726492c814b3644b7d376029d0). **Verify using the pinned commit hash (recommended):** A commit hash is cryptographically immutable, so this is the strongest way to ensure you are using the original signing key: ```bash cosign verify \ --key https://raw.githubusercontent.com/BerriAI/litellm/0112e53046018d726492c814b3644b7d376029d0/cosign.pub \ ghcr.io/berriai/litellm:v1.92.0 ``` **Verify using the release tag (convenience):** Tags are protected in this repository and resolve to the same key. This option is easier to read but relies on tag protection rules: ```bash cosign verify \ --key https://raw.githubusercontent.com/BerriAI/litellm/v1.92.0/cosign.pub \ ghcr.io/berriai/litellm:v1.92.0 ``` Expected output: ``` The following checks were performed on each of these signatures: - The cosign claims were validated - The signatures were verified against the specified public key ``` --- ## What's Changed * fix(proxy): gate team allowed_passthrough_routes to proxy admins by @ryan-crabbe-berri in https://github.com/BerriAI/litellm/pull/28097 * fix(tests): stabilize image-edit VCR cassettes to stop live gpt-image-1 spend by @mateo-berri in https://github.com/BerriAI/litellm/pull/28110 * fix(bedrock/cohere): send embedding_types as JSON array, not string by @ishaan-berri in https://github.com/BerriAI/litellm/pull/28172 * fix(tests): migrate realtime + rerank tests off shut-down upstream models by @yuneng-berri in https://github.com/BerriAI/litellm/pull/28191 * fix(caching): replay openai/responses bridge cache hits as chat streams by @Sameerlite in https://github.com/BerriAI/litellm/pull/28158 * Litellm oss staging by @Sameerlite in https://github.com/BerriAI/litellm/pull/28161 * feat(prometheus): add user_email and user_alias to user budget metrics by @Sameerlite in https://github.com/BerriAI/litellm/pull/28155 * test(callbacks): harden flaky proxy callback-leak detector by @yuneng-berri in https://github.com/BerriAI/litellm/pull/28195 * fix(bedrock): sanitize batch metadata to prevent Pydantic ValidationError by @mateo-berri in https://github.com/BerriAI/litellm/pull/28202 * fix(deepseek): use native /anthropic/v1/messages endpoint and sanitize tools by @mateo-berri in https://github.com/BerriAI/litellm/pull/28200 * feat(ui): add Interactions API endpoint to playground with SSE streaming by @Sameerlite in https://github.com/BerriAI/litellm/pull/28156 * fix(proxy): decode bytes and pass-through SSE for Google-native streamGenerateContent (#27444) by @Sameerlite in https://github.com/BerriAI/litellm/pull/28213 * refactor(bedrock/sagemaker): switch to lazy loading for response stre… by @harish-berri in https://github.com/BerriAI/litellm/pull/28189 * [Refactor] UI - Spend Logs: consolidate filter state and

## What's Changed - MCP server for CLI 3.0: shared MCP runner, `max_dim` guard. - Fix navigation readiness detection: per-target lifecycle event storage, and handle same-document navigations without burning the readiness timeout. - Fix markdown extraction destroying URLs and dropping long link lines. - Detect LLM output truncation instead of failing with misleading parse errors, and let truncation errors switch to the fallback LLM. - Keep plain BrowserErrors recoverable and preserve structured error memory through execute_action. - Fix visibility check mutating shared snapshot bounds. - Pin Browser Harness to `0.1.5` (Windows reliability fixes, cloud-browser suggestion). - CLI polish: shim for legacy CLI commands, hints for unknown commands, faster startup. - Restructure README quickstart around the CLI and Python library paths. **Full Changelog**: https://github.com/browser-use/browser-use/compare/0.13.3...0.13.4

## What's Changed - New interactive agent experience: running `ollama` now launches an agent to help you code and delegate work ``` ❯ ollama Ollama 0.32.0 ▸ Chat, Code, & Work (glm-5.2:cloud) Chat with models, code, search the web, and delegate real work ``` - Renamed the Codex App integration to ChatGPT: use ollama launch chatgpt (and --restore to return to your usual ChatGPT profile) - Simplified integration selection: the ollama launch menu now only offers the most popular integrations (other integrations can be accessed through `ollama launch` - Warns before launching older agent models: CodeLlama, Qwen2.5(-coder), Llama 3.x, Mistral, StarCoder, and the base DeepSeek-R1 tags now prompt a deprecation warning before ollama launch continues **Full Changelog**: https://github.com/ollama/ollama/compare/v0.31.2...v0.32.0