{"source":"ai-releases","note":"Free 5-item preview. The full feed is $0.05/call at GET /v1/ai-releases/changes.","question":"Did a tracked AI/ML framework & tooling tool ship a new release since T, and does it look breaking?","sample":[{"source":"ai-releases","entityId":"repo:openai/openai-python","type":"new_release","title":"openai/openai-python v2.46.0","summary":"## 2.46.0 (2026-07-17) Full Changelog: [v2.45.0...v2.46.0](https://github.com/openai/openai-python/compare/v2.45.0...v2.46.0) ### Features * **api:** /organization/projects/{project_id}/service_accounts/{service_account_id}/api_keys\" endpoint ([5a00941](https://github.com/openai/openai-python/commit/5a0094194eac9c605c8ca84d47d1c5518f8e2131)) * **api:** add owner_project_access to APIKeyListParams ([f589d04](https://github.com/openai/openai-python/commit/f589d04bf9f377ecb1f54335ab3ab9d825b5dfee)…","significance":6,"detectedAt":"2026-07-18T23:42:59.037Z","effectiveDate":"2026-07-17T02:47:14Z","sourceUrl":"https://github.com/openai/openai-python/releases/tag/v2.46.0","detail":{"repo":"openai/openai-python","version":"v2.46.0","prerelease":false,"breaking":false}},{"source":"ai-releases","entityId":"repo:ggml-org/llama.cpp","type":"new_release","title":"ggml-org/llama.cpp b10063","summary":"<details open> sync : ggml </details> **Website:** - <https://llama.app> **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b10063/llama-b10063-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/b10063/llama-b10063-bin-macos-x64.tar.gz) - [iOS XCFramework](https://github.com/ggml-org/ll…","significance":6,"detectedAt":"2026-07-18T23:42:59.036Z","effectiveDate":"2026-07-17T14:58:33Z","sourceUrl":"https://github.com/ggml-org/llama.cpp/releases/tag/b10063","detail":{"repo":"ggml-org/llama.cpp","version":"b10063","prerelease":false,"breaking":false}},{"source":"ai-releases","entityId":"repo:ggml-org/llama.cpp","type":"new_release","title":"ggml-org/llama.cpp b10064","summary":"<details open> opencl: transpose q4_K noshuffle scales for coalesced reads (#25805) </details> **Website:** - <https://llama.app> **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b10064/llama-b10064-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/b10064/llama-b10064-bin-macos-x64.t…","significance":6,"detectedAt":"2026-07-18T23:42:59.036Z","effectiveDate":"2026-07-17T15:40:55Z","sourceUrl":"https://github.com/ggml-org/llama.cpp/releases/tag/b10064","detail":{"repo":"ggml-org/llama.cpp","version":"b10064","prerelease":false,"breaking":false}},{"source":"ai-releases","entityId":"repo:ggml-org/llama.cpp","type":"new_release","title":"ggml-org/llama.cpp b10066","summary":"<details open> opencl: load and use `kernel_gemm_moe_q6_k_f32_ns` from bin kernel lib (#25797) </details> **Website:** - <https://llama.app> **macOS/iOS:** - [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b10066/llama-b10066-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/b10066/llama-b10066-bin-…","significance":6,"detectedAt":"2026-07-18T23:42:59.036Z","effectiveDate":"2026-07-17T23:13:36Z","sourceUrl":"https://github.com/ggml-org/llama.cpp/releases/tag/b10066","detail":{"repo":"ggml-org/llama.cpp","version":"b10066","prerelease":false,"breaking":false}},{"source":"ai-releases","entityId":"repo:ggml-org/llama.cpp","type":"new_release","title":"ggml-org/llama.cpp b10067","summary":"<details open> llama-quant : exclude i32 ffn_gate_tid2eid routing table from quantization (#25787) DeepSeek-V4's ffn_gate_tid2eid tensor is an i32 token-id -> expert-id index table, not weights. It was never added to the name-based exclusion list alongside ffn_gate_inp.weight, so llama-quantize tries to quantize it and fails since i32 cannot convert to a float type. Fixes ggml-org/llama.cpp#25754 Signed-off-by: Yash Raj Pandey <yashpn62@gmail.com> </details> **Website:** - <https://llama.app> *…","significance":6,"detectedAt":"2026-07-18T23:42:59.036Z","effectiveDate":"2026-07-18T12:12:58Z","sourceUrl":"https://github.com/ggml-org/llama.cpp/releases/tag/b10067","detail":{"repo":"ggml-org/llama.cpp","version":"b10067","prerelease":false,"breaking":false}}]}