Knowledge catalogue
llama: share the nextn tensor flags between models Follow-up of the TODO in glm5-next: move the trunk-only and MTP-only detection that each model copied into a nextn_flags helper of llama_model_base.
Explore how the GitHub Copilot agentic harness delivers strong results across multiple benchmarks and leading token efficiency, while maintaining flexibility to choose among more than 20 models. The p
CUDA: MMVQ nwarps=8 for bs=1 for dense models on DGX Spark Signed-off-by: ynankani ynankani@nvidia.com skip moe experts and allow others based on k geometry (allow only small idle tail) Signed-off-by:
ModelOpt checkpoints apply a float32 global scale to every projection output on top of the per-group quantization scales. Running the multiply and the cast back to the activation dtype as separate eag
What's Changed Models that don't set a repeat_penalty now default to 1.0 (off) instead of 1.1, matching other engines and speeding up speculative decoding; set a per-model parameter if an older model
Overview llama.cpp v0.6.0 introduces the new llama_batch_ext extended batch API (with llama_process) for mixed token/embedding inputs and MTP/deepstack state embeddings, adds support for the GLM-5.3-F
llama : fix unexpected graph reallocation in the k-pool models (#29958) llama : fix unexpected graph reallocation in the k-pool models Both k-pool models built a graph shape that depends on state the
model: add Qwen3.8-Flash-Next (qwen4exp) (#27742) gguf: add qwen4exp (Qwen3.8-Flash-Next) arch and converter Adds the GGUF-side plumbing for HF model_type qwen4_exp: MODEL_ARCH.QWEN4EXP plus tensors f
OpenVINO: Qwen3.5, memory optimization, and test-recurrent-state-rollback (#26952) OpenVINO backend: 1) enable gpt-oss moe on OV bk; 2) enable mxfp4 support OpenVINO backend: disable TOPK_MOE op test
test-save-load-state : print a per-model results table in --models mode (#29316) test-save-load-state : print a per-model results table in --models mode in --models mode the output was very heavy: eve
model : add Ling 3.0 VL support (#29151) model : fold Ling 3.0 VL into the BailingMoeV3 architecture Assisted-by: Scout model : keep shared NORM rope list intact when gating bailingmoe3 on mrope secti
Overview llama.cpp 0.4.1 adds Maple 20B-A1B, Tencent Hy 4, and Spark2.5 support, improves JSON schema handling, chat parsing, logging, and server child-process management, and updates ggml to v0.24.0.
model: add Kimi-K3 text model (#26185) model: add Kimi-K3 text model Hybrid KDA (linear) + MLA (full) attention as in Kimi-Linear-48B, plus five things that architecture does not have: cross-layer res
model : add support for MiniMaxText01ForCausalLM and MiniMaxM1ForCausalLM (#27018) llama : support for MiniMax-Text-01 model chore : renames to match the other MiniMax models model : add logits mask a
common: support --models-dir loading MTP assistant models (#24431) common: support --models-dir loading MTP assistant models common: preset: check for MTP models with strict prefix common: preset: Tak
model : add K2 Horizon dense and MoVA support (#29535) model: K2 Horizon gguf conversion code model: loading hparams and tensors in k2-horizon.cpp model: K2 Horizon compute graph model: K2 Horizon com
models : consolidate nextn row cropping into shared helpers (#30017) mimo2 : always emit h_nextn the other nextn-capable models set it unconditionally Assisted-by: pi:llama.cpp/MiMo-V2.6-Flash-MOPD mo
Overview This release focuses on backend performance and correctness, broader model coverage, and more robust server/router operation. It adds HRM-Text (DFM Mimir 1B) support, MiMo-V2.6 and HunyuanOCR
server: Dedup the draft HF model via dedup-cache-models (#27934) server: Dedup the draft HF model via dedup-cache-models Fixes #27846 server: avoid capturing structured binding in lambda Website: http
tests: add model resolution test on synthetic repo listings (#26172) tests: add model resolution test on synthetic repo listings Include download.cpp and arg.cpp inside a namespace with hf_cache monke
mtmd: support Qwen3-TTS (note: breaking change to llama-tts binary) (#26254) convert text model main model load ok convert encoder ok speaker encoder loading ok speaker enc graph adapt vocab for backb
tests : refactor test-recurrent-state-rollback (#29426) tests : use llama_context_ptr in test-recurrent-state-rollback Replace raw llama_context pointers with llama_context_ptr and drop the manual lla
models: pad on the left with ggml_pad_ext (#29567) models: pad on the left with ggml_pad_ext The Parakeet, LFM2-Audio, Granite Speech and Gemma 4 audio encoders build a left padding as a right pad fol
model: add MTP support for Nemotron model (#26725) model: add MTP support for Nemotron Nano model model: add mtp_flags for nemotron model address review comments Website: https://llama.app macOS/iOS:
model : Granite-Switch Architecture (#25107) granite-switch: add llama.cpp backend (POC, CPU) New 'granite-switch' architecture: a dense, all-attention Granite-4.1 model with N embedded LoRA adapters
Overview llama.cpp 0.4.0 adds initial Qwen3.8-Flash-Next and Nemotron-3-Puzzle support, on-demand tensor reading, per-slot server context limits, video input options, and a ggml update to 0.23.0 with
In controlled offline evaluations, HydraFusion’s selective coding workflows matched or exceeded the evaluated Opus 5 baseline while reducing estimated workflow cost. Now available as a research previe
src : add n_expert_used_max function (#28323) src : add n_expert_used_max function With Commit c61b98b ('model: add NVIDIA Nemotron-3-Puzzle-75B-A9B (NemotronHPuzzle) support (#25444)') it is now poss
Decision models Ollama now supports decision models through /v1/systemone, based on TypeSafe’s Jev API. Decision models return choices, probabilities, and scores instead of text. Use them for tasks su
metal : single-source fusion table + fusion debug rework (#28164) metal : rework fusion patterns into a single table All fusable op patterns for the Metal backend are now declared once in a fusion tab
model : fix MTP context kv cache allocation for deepseek2, glm4moe, … (#28630) model : fix MTP context kv cache allocation for deepseek2, glm4moe, cohere2moe architectures (#28626) model: add inverse
server + ui: fix stream routes for model names containing a slash (#26137) server + ui: refactor resumable stream routes to query string conv_id The conversation id can embed a model name containing s
spec : support DFlash for HunyuanOCR (#28890) model : add DFlash layer-input taps for HunyuanVL DFlash speculative decoding needs the target graph to expose the residual stream entering each layer (re
server: fix router eviction races with the existing queue (#29217) server: route every model load through the queue A model loaded by the fast path has no queue entry, so tick() evicts it at its LOADE
model : add support for HrmTextForCausalLM (DFM Mimir 1B) (#27625) model : add support for HrmTextForCausalLM (DFM Mimir 1B) HRM-Text runs two transformer stacks (low, high) in an alternating cycle ov
model is one package with three jobs: the contract between the runner and the architectures, the opened checkpoint, and building nn layers from checkpoint tensors. Its files did not say which was whic
model : fix SWA not being enabled for EXAONE 4.5 (#26848) model : fix SWA not being enabled for EXAONE 4.5 load_arch_hparams tests hparams.n_layer() == 64 before LLM_KV_NEXTN_PREDICT_LAYERS has been r
NVIDIA Nemotron 3.5 Lightning NVIDIA Nemotron 3.5 Lightning is an open 30B mixture-of-experts (MoE) model with 3B active parameters built for that execution layer of always-on agents. It is designed f
model : add LiquidAI/d1-3B decision model (#30110) model : add LiquidAI/d1-3b decision model mtmd : read LFM2 image resize algo from GGUF Assisted-by: Claude Opus 5.5 common : rename decision type d1
model : add LiquidAI/d1-omni-600M decision model (#30114) model : add LiquidAI/d1-omni-600M decision model Assisted-by: Claude Opus 5.5 mtmd : keep conformer GLU sigmoid on CUDA Assisted-by: Claude Op
Overview llama.cpp 0.3.0 introduces the dots3-note multimodal model (with a new DSA-ISWA KV cache), MTP support for GLM-4.5-Air, and tensor-split (-sm tensor) plus multi-sequence rollback fixes for De
model : GraniteSWAForCausalLM / GraniteMoeSWAForCausalLM (#25505) feat(convert): Add conversion for GraniteSWAForCausalLM Branch: GraniteSWAForCausalLM AI-usage: full (Bob, OpenCode + Qwen3.6-35b) Sig
tests : consolidate context state tests into test-llama-context (#30023) tests : consolidate the context state tests into test-llama-context test-save-load-state, test-recurrent-state-rollback and tes
models : remove redundant ggml_cont calls now supported by all unary kernels (#30198) models: remove redundant ggml_cont calls now supported by CUDA unary kernels models: restore cont_2d after sigmoid
GitHub’s context-aware secret classifier improves detection quality, raises weighted acceptance by 11.5%, and reduces median scan execution time by 86%. The post GitHub’s model for secret detection: F
ggml-openvino: update to 2026.4.1, optimize performance, expand ops, improve device listing. (#29852) ggml-openvino : Qwen3.5 MoE perf (#312) Squash of ravi9#312: ggml-openvino: add detailed inference
model : support classifier_pooling for rerankers (#29627) model : support classifier_pooling for ModernBERT rerankers Assisted-by: Claude Opus 5.5 model : read classifier pooling type in load_hparams
add GLM-5.3-Flash (GLM5-Next) support (#27773) Rebase GLM-Next support onto master, and migrate to llama-memory-hybrid-idx Add initial MTP support Merge branch optimizations. Reduce allocated compute
model : use exact GELU for ModernBERT encoders (#30108) model : use exact GELU for ModernBERT encoders Assisted-by: Codex model : keep tanh GELU aliases on ggml_geglu Assisted-by: Claude Opus 5.5 mode
server : improve startup log messages (#29125) server-models : show source per model in log Show [source] tag (preset/models_dir/cache) per model instead of cryptic * marker Show HF hub cache path in
model: fix all granite family parameter counts (#28643) model: fix all granite family parameter counts Signed-off-by: Aaron Teo aaron.teo1@ibm.com model: fix additional include, add missing A prefix f
model: support vllm decision 2 model (#30378) models: support pplx-decider vllm d2 update clarify naming rule stricter validation fix ci rm redundant filter_tensors Website: https://llama.app Attestat
model: support vllm d3 models (#30383) models: support pplx-decider vllm d2 update clarify naming rule stricter validation d3 Website: https://llama.app Attestations: https://github.com/ggml-org/llama
What's Changed llama: update llama.cpp to b11521 for decision models and EmbeddingGemma 2 by @jmorganca in #18917 server: reuse verified manifests and loaded models in inference routes by @dhiltgen in
llama, server: add /v1/systemone API (models: laya, julia-1, lev, openjev, kev) (#29818) init conversion convert: ok model loaded add server code improve conversion script support shared prompt prefix
OpenVINO: optimize stateful decode and GPU MoE inference (#28638) exclude GPU/NPU failing POOL_2D case Fix pool case ggml-openvino: fix stateful decode for Gemma-4 per-layer-type head sizes ggml-openv
Model-Saver: Write the SWA pattern, 15 more architectures roundtrip (#29042) llama: read the SWA pattern as a period or a per-layer array Add llama_model_base::load_swa_pattern(), which reads sliding_
tests : run test-save-load-state across all architectures (#27755) tests : run test-save-load-state across all architectures test-save-load-state previously only ran in ctest against a single download
Overview New version has been released. Nightly build: b10566 Web UI: the nightly-tag.txt asset contains the tag of the corresponding nightly release More info: dist : releases and versioning of ggml-
What's Changed Improved Gemma 4 tool calling and multi-turn reasoning, including more reliable tool-response continuations Fixed a recurrent MLX model cache leak that could increase memory use across