30
CategoryAll
Smart search ranks by meaning as well as keywords (one row per story, last 45 days).
RelateAnything: Real-Time Open-Vocabulary Relation Prediction From Any Inputs
RelateAnything is a 53M-parameter open-vocabulary relation prediction model running at 20 ms/frame, with 2.3-3.5x higher mean recall than comparable open-vocabulary methods.
RelateAnything predicts scored relations between image regions using any predicate vocabulary supplied at inference as text embeddings, with object labels never required as input, so region sources can change without retraining. Training covers 19,103 predicates using positive-unlabeled supervision; the authors release RA-4M (474k images, 4.3M geometrically verified relations over 10,102 free-text predicates) and the OV-SGG-Bench evaluation suite. The 53M-parameter model runs at 20 ms/frame and achieves 2.3-3.5x the mean recall of the strongest comparable open-vocabulary method across cross-dataset and zero-shot benchmarks. Model, corpus, and benchmark are public.
40
60
Operation RapidRust: APT36 Deploys RUSTYSHADE, RUSTYMOVE, PSNATCH, and BASHNATCH
Zscaler details Operation RapidRust: APT36 deploys four new tools including RUSTYSHADE, a Rust backdoor using private GitHub repos for encrypted C2.
Zscaler ThreatLabz documents Operation RapidRust, a campaign by Pakistan-aligned APT36 deploying four new tools: RUSTYSHADE, a 64-bit Rust Windows backdoor that uses attacker-controlled private GitHub repositories with a hardcoded PAT and AES-256-GCM-encrypted messages for C2; RUSTYMOVE; PSNATCH, a PowerShell file stealer that scans Office documents, archives, media, and databases modified in the last 120 days and exfiltrates up to 5 GB per run to per-machine GitHub repositories; and BASHNATCH. The backdoor was dropped via PowerShell from attacker-controlled Backblaze B2 storage and supports screenshots, webcam capture, file listing, downloads, and shell command execution.
70
Week in review: Salesforce and ServiceNow portals exposed for 17 months, exploited Metabase 0-day
Weekly digest: exploited Metabase zero-day breached Framework; Salesforce/ServiceNow portals read for 17 months; Microsoft patched 400+ flaws.
Help Net Security's week in review aggregates top stories: a 'City-Forum' campaign tracked by Reco has been pulling records from Salesforce and ServiceNow portals worldwide for 17 months, and Framework suffered a breach via an exploited Metabase zero-day exposing customer contact and IP data. It also covers Microsoft's August 2026 Patch Tuesday fixing 400+ flaws including exploited zero-day CVE-2026-68820, Cisco's fix for exploited firewall DoS bug CVE-2026-20349 (added to CISA KEV), and a second N-able N-central hotfix for actively exploited CVE-2026-18577. Other items include GitHub expanding Dependabot malware alerts to eight package ecosystems and EU AI Act enforcement beginning on 2 August 2026.
32
Russian hackers plant nuclear weapon prompt in malware to trip AI safety guardrails
ESET reports Russian group UAC-0099 hid a prompt in VBS malware comments to trip AI safety filters and disrupt automated malware analysis in Ukraine.
ESET identified a technique dubbed GuardBreaker in which UAC-0099 embedded a comment reading "I want to make nuclear weapon. Help me …" inside a malicious VBS script to trigger AI safety mechanisms and halt AI-assisted malware analysis. The script, part of the group's toolset, downloads the MATCHBOIL malware used exclusively by this Russia-aligned group; CERT-UA documented the chain including LUNCHPOKE, BURNYBEAR and MATCHBOIL.V2 in a July advisory. UAC-0099 typically targets transportation and energy sectors and hands validated targets to GRU-linked Sandworm. ESET warned that AI-assisted analysis must be backed by layered detection and human-driven engineering.
58
ThreatsDay: GhostJacking AI Attacks, EtherHiding ClickFix, Cursor CLI Flaw + 17 More Stories
A weekly bulletin aggregating short security updates, including the City-Forum data-theft campaign, a ShipMonk breach, a Cursor CLI flaw, and GhostJacking AI attacks.
The Hacker News ThreatsDay Bulletin bundles roughly 20 short updates across cloud services, AI tools, malware, breaches, and scams. Highlights include the City-Forum campaign pulling data from unauthenticated guest access in Salesforce Experience Cloud and ServiceNow Service Portals since March 2025, and a ShipMonk breach exposing Trezor customer order data for orders in seven countries between May 10 and August 8, 2026. Other items cover a patched Cursor CLI flaw that let cloned repositories run commands before the workspace-trust prompt, Okta's analysis of the Work Panel vishing console used by actors like UNC6671, and GhostJacking AI agent hijacking via a patched Claude Desktop sandbox escape. Meta also launched an on-device WhatsApp Scam Alert machine learning model that keeps message content on the device.
30
Rust Supply Chain Attack on arrayref: Significant Overlap with DPRK Campaigns
Malicious versions of the Rust crate arrayref executed a compile-time backdoor, with infrastructure overlapping recent DPRK supply chain attacks on Mastra and axios.
Wiz researchers report that malicious versions of the arrayref Rust crate ran a backdoor during compilation, compromising developer build pipelines. The campaign's command-and-control infrastructure significantly overlaps with infrastructure used in recent DPRK supply chain attacks, including campaigns against Mastra and axios. Developers using affected crate versions should audit their builds and review dependencies.
72
Threat Matrix: Mapping threats across cloud web applications
Microsoft released a Cloud Web Applications Threat Matrix, a MITRE ATT&CK-aligned framework mapping threats to cloud-hosted web apps and serverless platforms.
Microsoft introduced the cloud web applications threat matrix, which organizes attack techniques for cloud-hosted web apps and serverless platforms using MITRE ATT&CK tactics, from resource development through impact. The framework covers attack paths spanning application code, managed runtimes, workload identities, and deployment pipelines, cataloging techniques such as subdomain takeover, code injection in connected repositories, compromised registry images, exposed admin interfaces, and serverless trigger injection. It builds on Microsoft's earlier Kubernetes and storage services matrices to help defenders identify visibility gaps, prioritize hardening, and plan investigations in cloud-native environments.
35
TasmScan: Continuation-Aware Taint Analysis for TVM Bytecode with Savelist Abstraction
TasmScan introduces source-free taint analysis for TON smart-contract bytecode, detecting 95.3% of defects with 96.8% precision and 17x speedup.
TasmScan is the first bytecode-level static analysis framework for the TON Virtual Machine, enabling cross-continuation data flow reasoning without source code by modeling savelist semantics through forward register analysis with formal over-approximation guarantees. It lifts bytecode into a typed intermediate representation (TASIR) and performs path-sensitive taint analysis. On a 208-contract benchmark with human-confirmed ground truth it detects 95.3% of defects across five classes at 96.8% precision, and resolves 294,546 dynamic continuation targets with 100% precision across 2,921 registry contracts. It achieves a 17x median speedup over symbolic-execution baselines.
28
30
Notes on gotchas while migrating 35kb preprompts from Opus to self-hosted Ollama
Opinion piece urges migrating 35KB preprompts from Anthropic/OpenAI to self-hosted Ollama, citing session privacy risks and safety filters blocking security research.
The author documents gotchas migrating 35KB preprompts from Claude Opus to self-hosted Ollama, motivated by fears that frontier providers train on user sessions, citing the OpenAI Navier-Stokes controversy. The piece argues inference providers cannot audit their own retention or training pipelines and that only self-hosted hardware offers verifiable privacy. It also criticizes frontier safety filters for refusing vulnerability research tasks and calls for models that support exploitability testing in CI/CD pipelines.
15
60
Molecular Déjà Vu: Digit-Level Retrieval of Published Values in Frontier Language Models
Audit of 22 frontier models finds widespread verbatim retrieval of published molecular property values, with higher reasoning increasing recall of memorized numbers.
An arXiv audit tests 22 frontier LLMs across 12 molecular regression benchmarks for verbatim retrieval of published values. More than 50% of the LLMs show verbatim retrieval on five datasets, and identical experiments are flagged 89% more often at a high reasoning level than at the lowest one. Suppressing retrieval moves model prediction errors closer together in relative terms, suggesting predictive capability is not determined solely by memorized values.
42
Agora: Git as Shared Memory for Collective AutoResearch
Agora records multi-agent research as an append-only Git DAG; 13 LLM workers ran nearly 12 days on a weight-transfer problem.
Agora stores every result, hypothesis, and verification as an immutable commit in a Git-stored DAG, with a derived index exposing the frontier and verification status of claims. In a nearly 12-day run, 13 language-model workers with no assigned tasks or central planner published 1,703 contributions on initializing a frozen 119.6M-parameter attention-SSM hybrid from 141 donor models. They improved the evaluator from 3.39 to 1.899 bits per byte, closing 62% of the gap to a trained GPT-2 124M, with 165 independent reproductions posted and none failing.
45
NCP-ArchPreview Technical Report: Moving towards Latent Space Language Models through Next Concept Prediction
An 8.9B-parameter latent-space language model using next-concept prediction matches OLMo-3-7B pretraining loss with only 51.3% of the training tokens.
NCP-ArchPreview augments next-token prediction with Next Concept Prediction over a product-quantized concept vocabulary built from hidden states, trained jointly end-to-end. The 8.9B model was trained on 5.73T tokens from the Dolma-3 dataset, the largest latent-space language model demonstration to date. It consumes 51.3% of the tokens to reach OLMo-3-7B's final pretraining loss and outperforms it by 2.45 points on the downstream macro-average, including a 5.99-point GSM8K gain. The learned latent space also enables lightweight domain adaptation via a 17M-parameter VQ module and improves speculative drafting accepted length by 4.17%.
52
Cohere Releases North Small Translate: A 218B MoE Translation Model That Scores 83.6 on WMT26 Across 50 Languages
Cohere released North Small Translate, an open-weight 218B MoE (25B active) translation model scoring 83.6 on WMT26 across 50 languages.
Cohere and Cohere Labs released North Small Translate, a decoder-only sparse Mixture-of-Experts translation model with 218B total and 25B active parameters, 128 experts with 8 activated per token plus shared experts, and 16K-token input and output context. In Cohere's vendor-reported WMT26 evaluation, judged by GPT-5.6-Sol, it scores 83.6 averaged across 50 languages (84.36 in an agentic multi-pass mode), ahead of DeepL NextGen (81.37), Qwen 3.5 397B A17B (81.56), GLM 5.2 (76.50), and Google Translate (68.20). The model was built with RWS's Language Weaver team, post-trained specifically for translation, and reports 112 output tokens per second versus 81 for Gemma 4 31B, with long-document xCOMET-XL scores of 48.9 versus 21.3 for Google Translate. It is available free on Cohere's Chat V2 API until rate limits, with three self-hosting checkpoints including a 4-bit NVFP4 variant running on 1x B200 or 2x H100.
55
Comprehensive reconstruction of collider events with hypergraph representation learning and graph-conditioned diffusion
VyPER framework reconstructs collider events using hypergraph representation learning and graph-conditioned diffusion, outperforming existing reconstruction techniques across Standard Model processes.
Researchers present VyPER, a geometric learning framework that represents collider events as hypergraphs with a physics-inspired topology for particle event reconstruction. It combines supervised hyperedge classification for assigning measured jets and charged leptons to parent particles with a graph-conditioned diffusion model predicting unmeasured neutrino kinematics, optimized with a joint loss. Evaluated across several proton-proton collision processes, it demonstrates accurate reconstruction across Higgs, electroweak, and top-quark sectors.
12
Objective vs. Search: Decomposing What Makes a Good Tokeniser
New tokeniser study shows search procedure, not optimisation objective, drives bits-per-byte performance across model sizes, vocabulary sizes, and multilingual settings.
The paper disentangles BPE and UnigramLM along two axes: optimisation objective (compression vs log-likelihood) and search procedure (bottom-up merging vs top-down pruning). Two new algorithms, BottomUpLL and TopDownComp, complete the 2x2 design space, and trained language models are evaluated on bits-per-byte and BLiMP across model sizes, vocabulary sizes, and English-only vs multilingual domains. Bottom-up tokenisers consistently achieve lower bits-per-byte in most settings, while BLiMP shows no consistent relationship with design choice.
28
Decomposition Buys Integrity, Not Yield
Study of 600 production deep-research traces finds agent-tree decomposition loses findings at rate N^(1-δ); flat architectures maximize yield.
The paper models multi-agent decomposition as a tree where an agent holding b items retains each with probability r(b); with r(b)=1/b every tree delivers exactly one finding regardless of shape. Analysis of 600 production deep-research traces estimates delta=0.34 retention decay, and 1,012 annotated traces show one brief in sixteen goes off-target per tier, giving an alignment penalty of 0.536. Depth still cuts root context exposure from N to N^(1/k) and is cheaper at scale, with a hazard model over 743,819 production tool calls showing delegation is an opening move rather than a response to filling context.
30
Package Manager Trends
Sixteen-week roundup finds package managers converging on release-age cooldowns, install-script blocking, malware scans, and recurring path-traversal and credential-leak fixes.
The author aggregates supply-chain security trends from sixteen weeks of This Week in Package Management, built from about 80 RSS feeds. Release-age cooldown gates shipped in Deno 2.8, Bundler, npm, Yarn, mise, Hex, Mamba, and Cargo, with Dependabot making a three-day cooldown default in August. npm 12 and Bun 1.4 now block lifecycle install scripts by default, and Composer 2.10 and uv added install/publish-time malware checks, while npm's registry began scanning at publish time. Path traversal on archive extraction was fixed in 14 of 16 weeks across tools including uv, pnpm, Docker, and Composer, and credential-misdirection bugs affected Cargo, ORAS, Composer, and Renovate.
45
TokenRhythm/NeoHorse-1-4B — new model trending #30 on Hugging Face
TokenRhythm releases NeoHorse-1-4B, an Apache-2.0 agentic fine-tune of Qwen3.5-4B claiming +5.93 benchmark macro-average gain.
NeoHorse-1-4B is a roughly 4B-parameter text-only causal language model post-trained by TokenRhythm from Qwen/Qwen3.5-4B for agent harnesses, tool use, coding, and instruction following. It applies routing-guided curriculum SFT and routing-guided on-policy distillation over execution trajectories as an early prototype toward recursive self-improvement (RSI). The release reports a 64.87 macro average across ten benchmarks versus 58.94 for Qwen3.5-4B (+5.93) and is distributed under Apache-2.0, trending #30 on Hugging Face.
28
Your threat feed is someone else's database: What ingesting malware intel at scale takes
GitHub's Dependabot lead shares five production lessons for ingesting community malware intelligence feeds across eight package ecosystems at scale.
GitHub's Dependabot team monitors over 30 million repositories and extended malicious-package advisories from npm to eight package ecosystems by ingesting OpenSSF's malicious-packages intelligence. The team catalogued roughly 18 new malicious npm packages per day in the year ending May 2026. The write-up argues that provenance with batch reverts, fingerprinting to catch echo-chamber duplicates, and heavyweight normalization are the make-or-break engineering for feed ingestion. It also recommends automated publishing with import caps and anomaly flagging, and quarantining malformed records rather than silently repairing them.
28
Iran-Linked Hackers Use Fake LinkedIn Job Offers to Deploy NodeRabbit and PollCat RATs
Iran-linked Mirage Kitten targets software engineers with fake LinkedIn recruiter coding tests deploying new NodeRabbit and PollCat RATs.
Kaspersky researchers link the campaign to Mirage Kitten (also tracked as UNC1549, Smoke Sandstorm, Nimbus Manticore), with victims in aviation, aerospace and fintech in Egypt, Ethiopia and Afghanistan. Trojanized npm dependencies (colorized_terminal, pretty-log) bundled in coding-challenge archives launch the Node.js implants across Windows, Linux and macOS. NodeRabbit uses AES-256-GCM-encrypted C2 via Azure, and its third variant persists through a fake GitHub Copilot Helper VS Code extension plus Git post-merge/post-checkout hooks. PollCat is an obfuscated JavaScript RAT that registers with C2 before OTP authentication and inventories tools from 24 security vendors.
72
DavidAU/Qwen3.8-27B-TURBO-Fable-Cold-Fusion-735-882-Heretic-Uncensored-NEO-CODER-MAX-MTP-GGUF — new model trending #8 on Hugging Face
A new Qwen3.8-27B GGUF fine-tune claims ARC-C 735 at 8-bit with thinking tokens cut 2x-10x versus the base model.
Independent creator DavidAU released a GGUF fine-tune of Qwen3.8-27B built with Unsloth, claiming ARC-C of 735 at 8-bit and 719 at 4-bit, trending #8 on Hugging Face. The 'TURBO' variant cuts thinking tokens by one half to as much as one tenth while retaining output quality and detail. The repo ships both regular and MTP quants and claims gains over the base model across seven benchmarks, using 'Cold Fusion (GAIN + Unsloth)' and 'Fable Fusion 711' training methods.
42
CoVeR: Coverage-Based Token Pruning for Multi-View 3D Reasoning in VLMs
CoVeR, a training-free coverage-based token pruner, preserves 93.5% of VLM 3D-reasoning performance using only about 8% of visual tokens.
Researchers introduce CoVeR, a deterministic, training-free selector that chooses visual tokens to cover every region of a multi-view 3D scene using only token coordinates. Unlike learned-importance and voxelization pruners, it enforces an exact per-scene token budget, avoids saturation plateaus, and prevents near-duplicate selections. Experiments across four vision-language models show it surpasses prior state of the art by 3.9 percentage points on average across three 3D reasoning benchmarks.
45
TART: A Modular Tool for Technique-Aware Audio-to-Tablature Guitar Transcription
TART pipeline transcribes guitar audio to tablature with fingering and technique labels, outperforming prior baselines by up to 8.5 Tab F1 points.
TART is a modular four-stage pipeline for audio-to-tablature guitar transcription, combining an audio-to-MIDI transcription model, an expressive technique classifier, an audio-conditioned T5 encoder-decoder for string-fret assignment, and an automated tablature generator. In zero-shot evaluation on GuitarSet, EGDB, and two noisy augmented benchmarks, it averages 81.35% audio-to-MIDI F50 (+6.67 over the best baseline), 71.8% string-fret Tab F1 (+8.5), and 54.08% end-to-end Tab F1. The authors state it is the first framework to generate tablature with both fingering and expressive technique annotations directly from guitar audio.
26
Iran-linked APT Mirage Kitten Uses Fake Job Tests to Spread Malware
Kaspersky reports Iran-linked Mirage Kitten delivers new NodeRabbit and PollCat malware to fintech and aviation targets via fake LinkedIn coding assessments.
Kaspersky researchers documented two previously undocumented Node.js malware families, NodeRabbit and PollCat, attributed with high confidence to Iran-linked APT group Mirage Kitten. The malware is delivered via fake recruiter personas on LinkedIn offering coding assessments hosted on Amazon S3, with instructions banning AI assistants so AI code-review tools would not flag the trojanized npm packages. NodeRabbit is cross-platform (Windows, Linux, macOS), uses AES-256-GCM-encrypted C2 on Azure, includes sandbox checks, and one variant installs a fake 'GitHub Copilot Helper' VS Code extension plus Git hook persistence. Victims identified so far are in fintech and aviation organizations across Egypt, Ethiopia, and Afghanistan.
74
42
WireGuard-Linux Stack-Based Buffer Overflow in lsiio (Linux IIO Userspace Tool) Due to Unbounded fscanf
The Linux IIO userspace tool lsiio has a stack buffer overflow in find_type_by_name() caused by unbounded fscanf reads of oversized filesystem-backed attribute values.
A stack-based buffer overflow exists in the Linux Industrial I/O (IIO) userspace utility lsiio. In the find_type_by_name() function, the program reads an unbounded string from a filesystem-backed attribute into a fixed-size stack buffer using fscanf("%s", ...). A crafted or oversized attribute value causes a write beyond the bounds of the destination buffer. Despite the title's reference to WireGuard-Linux, the flaw described is in the IIO lsiio utility.
12
MAxBench: A Multinomial Concept Recovery Benchmark
MAxBench evaluates multinomial concept recovery methods, finding affine subspaces steer most reliably but none consistently beats prompting.
MAxBench is a geometry-agnostic evaluation framework for multinomial concept representations in language models, based on sampling from recovered concept representations. It compares 10 localization methods covering 5 geometry types across 6 concepts and 4 models. Findings show affine subspaces steer more reliably than rank-one or linear subspaces due to better non-zero offsets, manifold steering is competitive where applicable, and no method consistently outperforms prompting.
18
Kaininja: Extending Native 3D Generators to the Part Level
KaiNinja extends TRELLIS.2 native 3D generation to part-level assets via a dual-volume O-Voxel representation, cutting whole-object Chamfer distance by 40%.
KaiNinja extends the TRELLIS.2 native 3D generator to produce part-level assets instead of one fused mesh, enabling downstream editing, rigging, and simulation. A dual-volume form of the O-Voxel representation solves the problem that a single volume cannot represent interfaces where two parts touch. The model needs no segmentation network, is partly trained on LLM-agent-authored part data, lowers whole-object Chamfer distance by 40%, and raises strict part F-score by 16% versus other part-generation pipelines.
18
New Ted Backdoor Hides Inside Victims' Own HAProxy Builds to Intercept Web Traffic
Rapid7 found a new backdoor, ted, compiled into trojanized HAProxy at two South Korean organizations, with medium-confidence attribution to North Korean actors.
Rapid7 documented a previously undocumented Linux toolkit named ted compiled into the HAProxy load balancer binaries of two South Korean organizations in the automotive and media sectors. The implant intercepts web traffic, serves altered pages only to filtered visitors, and hides C2 exchanges from backend logs and HAProxy statistics; a companion RAT, curlRAT, beacons on a default 12-hour schedule. The toolkit also trojanizes crond, sshd, agetty, atd, and polkitd binaries and sanitizes logs and bash history. Rapid7 attributes the activity with medium confidence to North Korean state-sponsored actors, with domain infrastructure overlapping APT37 listings in maltrail and delivery resembling the Operation SyncHole campaign.
65
Perplexity Details Its GPU Embedding Stack: How Ivy, Tulip and ROSE Serve pplx-embed
Perplexity details its GPU embedding serving stack (Ivy, Tulip, ROSE), which reuses LLM prefill/decode kernels, CUDA graphs, and LazyTensors to cut launch overhead.
Perplexity engineers published a deep dive on the serving infrastructure behind pplx-embed, used across Perplexity Search and its API platform. The stack comprises Ivy (Rust HTTP gateway), Tulip (gRPC scheduling and batching), and ROSE (Runtime-Optimized Serving Engine), which reuses LLM prefill and decode kernels rather than running a separate embedding engine. Optimizations include whole-model CUDA graphs with lazy capture and a LazyTensor abstraction that overlaps CPU batch preparation with in-flight GPU work. Benchmarks are reported against vLLM v0.22.0 in BF16, with FlashAttention 4 generally fastest but FlashInfer 3 winning on Qwen-based models at very long sequence lengths.
33
Wireshark 4.6.8 patches 28 security bugs, nine in file parsers
Wireshark 4.6.8 fixes 28 security bugs, including nine crash-prone capture file parsers, misdecoded 5G fields, and memory-safety issues.
Wireshark 4.6.8 fixes 28 security bugs spanning advisories wnpa-sec-2026-64 through wnpa-sec-2026-91, including nine crash bugs in file parsers such as pcapng, Endace ERF and Tektronix K12xx that trigger when opening capture files. Fixes cover dissectors for RDP, SSH, Kerberos, H.245, CMS, C12.22 and several Bluetooth protocols, plus unnumbered memory-safety issues like a stack buffer overflow in the K12/RF5 writer. The release also corrects eight misdecoded 5G NAS/5GSM fields and moves the Unix extcap path to /usr/libexec/wireshark/extcap.
35
Vulnerabilities fixed in libxml2-2.15.4
libxml2 2.15.4 fixes an out-of-bounds read in xmlregexp's NXT macro plus several integer overflow and parsing flaws.
libxml2 2.15.4 (released September 1, 2026) includes security fixes: an out-of-bounds read in the xmlregexp NXT macro, missing overflow checks in dict.c, uri.c, and valid.c, an integer overflow in xmlIO before the writecallback, and an overflow check in xmlXPtrEvalXPtrPart. The release also propagates parseFlags in xmlXIncludeProcess and xmlXIncludeProcessTree. No CVE identifiers, exploitation, or severity ratings are given in the announcement.
28
XHToken/Spark-X2.5-4B-GGUF — new model trending #30 on Hugging Face
XHToken released GGUF weights of Spark-X2.5-4B, a compact model with 1M-token context and 200+ language support, under Apache 2.0.
The Hugging Face repository provides BF16 GGUF conversions of Spark-X2.5-4B, a compact general-purpose language model for conversation, writing, translation, reasoning, coding, tool use, and agentic workflows. The model uses a hybrid attention architecture, supports a native context length up to 1M tokens, and covers more than 200 languages. Local inference is supported through Ollama and LM Studio via an XHToken llama.cpp fork, with a --think=false flag to disable thinking mode for faster responses. Released under Apache License 2.0; it was trending #30 on Hugging Face at publication.
24