18 this week·34 this month·76 all-time
Open-weight model releases and local inference tooling
Gemma 4 was launched by Google under an Apache 2.0 license, marking a significant open-model release focused on reasoning, agentic workflows, multimodality, and on-device use. It outperforms models 10x larger and has…
NVIDIA’s Nemotron 3 Super is a 120B parameter / ~12B active open model featuring a hybrid Mamba-Transformer / SSM Latent MoE architecture and 1M context window, delivering up to 2.2x faster inference than GPT-OSS-120B…
Release v5.8.0 New Model additions DeepSeek-V4 DeepSeek-V4 is the next-generation MoE (Mixture of Experts) language model from DeepSeek that introduces several architectural innovations over DeepSeek-V3. The…
vLLM v0.20.0 Highlights This release features 752 commits from 320 contributors (123 new)! DeepSeek V4: Initial DeepSeek V4 support landed (#40860), with DSML token-leakage fix in DSV4/3.2 (#40806), DSA + MTP IMA fix…
Alibaba released Qwen3.6-27B, a dense, Apache 2.0 open coding model with thinking and non-thinking modes, outperforming the larger Qwen3.5-397B-A17B on multiple coding benchmarks including SWE-bench and Terminal-Bench…
Moonshot's Kimi K2.6 is a major open-weight 1T-parameter MoE model featuring 32B active parameters, 384 experts, MLA attention, 256K context window, native multimodality, and INT4 quantization. It supports day-0…
vLLM v0.19.0 Highlights This release features 448 commits from 197 contributors (54 new)! Gemma 4 support: Full Google Gemma 4 architecture support including MoE, multimodal, reasoning, and tool-use capabilities…
Gemma 4 Effective 2B (E2B) ``` ollama run gemma4:e2b ``` Effective 4B (E4B) ``` ollama run gemma4:e4b ``` 26B (Mixture of Experts model with 4B active parameters) ``` ollama run gemma4:26b ``` 31B (Dense) ``` ollama…
Nimbus builds production AI systems — internal tools, customer agents, retrieval pipelines — combining humans and AI end-to-end. From scoped pilot to production in 4–8 weeks.
vLLM v0.16.0 Please note that this release was branch cut on Feb 8, so any features added to vLLM after that date is not included. Highlights This release features 440 commits from 203 contributors (7 new)! Async…
Highlights This release features approximately 660 commits from 251 contributors (86 new contributors). Breaking Changes: Async scheduling is now enabled by default - Users who experience issues can disable with…
MiniMax M2.1 launches as an open-source agent and coding Mixture-of-Experts (MoE) model with ~10B active / ~230B total parameters, claiming to outperform Gemini 3 Pro and Claude Sonnet 4.5, and supports local inference…
GLM-4.7 and MiniMax M2.1 open-weight model releases highlight day-0 ecosystem support, coding throughput, and agent workflows, with GLM-4.7 achieving a +9.5% improvement over GLM-4.6 and MiniMax M2.1 positioned as an…
vLLM v0.12.0 Release Notes Highlights Highlights This release features 474 commits from 213 contributors (57 new)！ Breaking Changes: This release includes PyTorch 2.9.0 upgrade (CUDA 12.9), V0 deprecations including…
Highlights This release includes 1456 commits from 449 contributors (184 new contributors)! Key changes include: PyTorch 2.9.0 + CUDA 12.9.1: Updated the default CUDA build to `torch==2.9.0+cu129`, enabling Inductor…
Highlights This release features 538 commits, 207 contributors (65 new contributors)! This release completes the removal of V0 engine. V0 engine code including AsyncLLMEngine, LLMEngine, MQLLMEngine, all attention…
DeepSeek-V4 makes million-token context a serving-systems problem. Together AI explores the inference work behind V4 on NVIDIA HGX B200, including compressed KV layouts, prefix caching, kernel maturity, and endpoint…
vLLM v0.20.1 This is a patch release on top of `v0.20.0` primarily focused on DeepSeek V4 stabilization and performance improvements, along with several important bug fixes. DeepSeek V4 Base model support (#41006)…
Release v5.7.0 New Model additions Laguna Laguna is Poolside's mixture-of-experts language model family that extends standard SwiGLU MoE transformers with two key innovations. It features per-layer head counts allowing…
Ollama is now powered by MLX on Apple Silicon in preview Ollama on Apple silicon is now built on top of Apple’s machine learning framework, MLX, to take advantage of its unified memory architecture…
vLLM v0.18.0 Known issues Degraded accuracy when serving Qwen3.5 with FP8 KV cache on B200 (#37618) If you previously ran into `CUBLAS_STATUS_INVALID_VALUE` and had to use a workaround in `v0.17.0`, you can reinstall…
vLLM v0.17.0 Known Issue: If you are on CUDA 12.9+ and encounter a `CUBLAS_STATUS_INVALID_VALUE` error, this is caused by a CUDA library mismatch. To resolve, try one of the following: 1. Remove the path to system CUDA…
Highlights This release features 335 commits from 158 contributors (39 new)! Model Support New architectures: Kimi-K2.5 (#33131), Molmo2 (#30997), Step3vl 10B (#32329), Step1 (#32511), GLM-Lite (#31386), Eagle2.5-8B…
South Korea's Ministry of Science launched a coordinated program with 5 companies to develop sovereign foundation models from scratch, featuring large-scale MoE architectures like SK Telecom A.X-K1 (519B total / 33B…
vLLM v0.13.0 Release Notes Highlights Highlights This release features 442 commits from 207 contributors (61 new contributors)! Breaking Changes: This release includes deprecation removals, PassConfig flag renames, and…
Nimbus builds production AI systems — internal tools, customer agents, retrieval pipelines — combining humans and AI end-to-end. From scoped pilot to production in 4–8 weeks.
NousResearch's Nomos 1 is a 30B open math model achieving a top Putnam score with only ~3B active parameters, enabling consumer Mac inference. AxiomProver also posts top Putnam results using ThinkyMachines' RL stack…
Highlights This release contains 740 commits from 266 contributors (97 new)! Breaking Changes: This release includes PyTorch 2.8.0 upgrade, V0 deprecations, and API changes - please review the changelog carefully…
opencl: add opt-in Adreno xmem F16xF32 GEMM for prefill (#22755) ggml-opencl: add Adreno xmem F16xF32 GEMM for prefill ggml-opencl: address Adreno xmem review comments ggml-opencl: align xmem gemm kernel naming…
CUDA: handle OW > 65535 in im2col (2D and 3D) (#22944) `im2col_cuda` and `im2col_3d_cuda` both dispatch with `block_nums.y = OW`. CUDA caps grid Y at 65535. Conv1d encoders on raw 16 kHz audio with T > 65535 (~ 4 s)…
spec : parallel drafting support (#22838) spec : refactor spec : drop support for incompatible vocabs spec : update common_speculative_init() cont : pass seq_id cont : dedup ctx_seq_rm_type server : sketch the ctx_dft…
Gemma 4 MTP (Multi-token Processing) for the MLX runner Gemma 4 MTP speculative decoding is now supported on Macs. This can give over a 2x speed increase for the Gemma 4 31B model on coding tasks. ``` ollama run…