Find published research by company, product, platform or technology.
Showing 1–20 of 79 dossiers
SQLite 3.54 is a compatibility release worth checking: standalone sqlite3_analyzer is deprecated, CLI behavior shifts, authorizer checks expand and Windows XP builds are no longer supported.
GLiNER2.5-Decide attacks the same bounded-decision layer as Jev and CLM from a much smaller encoder architecture. Its strongest benchmark claims are vendor-produced, but CPU deployment and constrained joint decoding make it a materially different option for software-facing AI decisions.
Meta’s Muse Glimmer 30B combines tool use, coding, vision and agentic task completion with official local-runtime artifacts. A 17GB GGUF build targets 24GB-VRAM machines, but Meta also attaches a separate usage policy, so builders should distinguish weight availability from unrestricted use.
Turso’s hosted early preview adds `BEGIN CONCURRENT` transactions backed by MVCC. Writes to different rows can proceed in parallel, while conflicting transactions fail at commit and must retry. The feature targets a core scaling constraint that often pushes applications away from SQLite-style architectures.
The important development is active exploitation, not the original vulnerability disclosure. WordPress operators running Elementor Pro 4.2.1 or earlier should treat this as an immediate patch-and-hunt event, especially on sites with public forms that include optional file uploads.
The change makes heavier frameworks and dependency trees deployable to Workers without plan-specific compressed-size ceilings, but it also changes what builders need to measure: the operative limit is now uncompressed Total Upload rather than the gzip number they may have optimized around.
AWS is changing how Lambda introduces managed runtimes: Node.js 26 and Python 3.15 are available in public preview before GA, with normal runtime identifiers that automatically graduate when the runtimes become production-ready.
The funding headline is less interesting than the workload signal: Supabase says agents now create most new databases on its platform, and it is buying Turso to handle higher-volume database creation for those workloads.
The counting-rule change is no longer theoretical. Early post-cutover data suggests public views can materially outpace Engaged views, with the size of the gap varying by channel size, category and discovery surface.
This was not a Firecracker escape or access to a live victim disk. It was a storage-isolation failure underneath the sandbox: researchers recovered foreign directory structures, database pages and complete SQLite databases from reused blocks, and Cloudflare had to fix allocation plus retire existing disks and cached snapshots.
DV360's new bulk-campaign file format isn't a drop-in CSV upgrade: targeting expands, YouTube vendor columns change, and API support lags the interface. Integrators should audit parsers before migrating.
The useful part is not the 800,000-line headline. GitHub has published unusually detailed receipts for a production-scale agent-assisted migration: roughly $120,000 of token spend, 14.5 weeks of incremental releases, dozens of regressions, extensive compatibility tests and a workload-specific jump from 7.55 to 120 session lifecycles per second.
This is a useful reminder that exploitation pressure does not scale neatly with plugin popularity: Wordfence says it has blocked more than 250,000 attempts against a plugin with a five-figure install base.
Teams with pinned, custom-image or auto-update-disabled GitHub Actions runners can now see registration or job execution fail before the September 25 cutoff. The migration is not just a one-time jump to v2.329.0: already-registered runners must also stay within 30 days of the latest runner release.
Project Zenith is not a new model or another Copilot feature. It standardizes a developer-focused Windows experience and hardware floor for local AI work, with preconfigured tooling and OS settings intended to reduce setup friction and dependence on metered cloud inference.
K2 Horizon is notable less for another benchmark claim than for reproducibility: IFM is publishing model weights, architecture, training code, data or construction recipes, evaluation resources and intermediate training material instead of stopping at a final checkpoint.
The median SaaS LTV forecast looks almost right at 12 months, but that average hides huge misses in both directions. For acquisition budgets, payback planning and company valuation, ChartMogul’s new 3,331-company analysis argues for treating LTV as a directional indicator rather than a precise revenue forecast.
GLM-5.3-Flash combines open weights, multimodal coding/agent capability and an 18B-active sparse architecture with a large anonymous pre-launch trial. Z.ai has already issued a chat-template correction for early downloads, showing that day-one self-hosted deployments need artifact-level validation as well as model benchmarking.
Sentence Transformers 6 now has both unified multi-vector inference and a documented end-to-end training workflow. A new project-authored benchmark shows fast domain adaptation on a single GPU, but the result is workload-specific and index costs remain high.
Groq 3 LPX is moving from architecture announcement to manufactured infrastructure. Artificial Analysis measured about 3,400 output tokens/s at both 10K and 100K context on an NVIDIA-hosted private endpoint, but the single-concurrency benchmark does not yet establish public-cloud price, multi-tenant throughput or end-to-end agent speed.