Topic
The shift by AI vendors and users toward compact, low-cost language models that are cheaper and easier to deploy than large general-purpose ones.
AWS open-sourced Strands Decider 2B, a 2B-parameter decision model it says makes a decision in a median of about 115 ms on an RTX 3090. Strands Box, the sandbox released with it, enforces agent rules in a kernel-level policy layer outside the model.
Reality
Harmonic Security raised a $50 million Series B led by Crosspoint Capital Partners, lifting its total funding to $76 million. Calcalist put rival Prompt Security's sale at about $250 million on $23 million raised, the nearest price comparison.
Perspective Coverage
6 publishersReality
Coinbase says a Qwen3.5-9B model post-trained for under $100 beat Claude Opus 4.5 on its Onramp fraud benchmark. The figures are Coinbase's on private data; the reusable parts are the label-based reward and the tier-only, block-only harness.
Reality
A 26M-parameter edit tagger scored F0.5 0.799 at phone-keyboard grammar correction against 0.663 for fine-tuned small LLMs, its authors report. It altered correct text 2.4% of the time against 20.1%, on a benchmark the authors ran themselves.
Reality
Microsoft released Microsoft-Decision-1, a small model that returns calibrated probabilities for fixed options, priced at $0.042 per million input tokens. The accuracy and speed figures are Microsoft's own, and the base model is due to change soon.
Reality
Anthropic launched Claude Haiku 5.5 at an average price about 75% below Haiku 4.5. The saving varies widely with prompt length, so teams moving classification, support or query traffic need to price their own requests before they switch models.
Perspective Coverage
22 publishersReality
Microsoft released Decision-1, a model that scores fixed answer options and, by its own benchmarks, runs 35 times faster than GPT-6 Sol. Whether routing work moves off general-purpose models now depends on independent tests and on pricing that confirm those self-reported figures.
Reality
Jeff, an open 0.8B model, returns a probability for each option in one forward pass and sends low-confidence agent decisions to Qwen3.8-27B. The confidence score attached to each answer tells an agent when a small decision is worth the larger model's time.
Reality
Qwen2.5-3B, wired to a local Wikipedia index, scored 52% on 150 post-cutoff questions it answers none of unaided, up from 33%, in a dev.to author's tests. Each fix targets a measured 3B failure, so a zero-shot 7B gained only 9 points from them, and the two readers' confidence intervals overlap.
Reality
Capsule Security has shipped an evaluator that runs inside an agent's execution path and can stop an intended action before it executes. Doing that requires adding a chokepoint to the agent's execution path.
Reality
The startup says Anchor 3.0 caught more than 90 percent of violations in a test it published itself, at under a 500th of the cost of a frontier call, and the messages it misses stay the firm's problem.
Reality
A case study reports an 8B model that got better at naming threat categories while losing ground on severity scoring, lifecycle counting and declining a weather question. Its author says every failure mode is already in the literature.
Reality
A Microsoft case study credits Docusign's swap to small task-specific models with 90% lower cost and eight times the throughput. Docusign's own accounting of the whole pipeline claims 50 times cheaper per document.
Reality
A KDnuggets walkthrough builds its comparison on Qwen2.5-0.5B-Instruct on an M2 MacBook Air, where the keys and values for a static instruction block come out identical on every call and only the ticket text changes.
Reality
ShadowPEFT is in Hugging Face PEFT's main branch as of a September 15th announcement, and it carries its own hidden state and can be detached as a smaller standalone model. Trying it means installing PEFT from source.
Reality
Microsoft's local runtime handles model download, hardware detection and execution provider selection behind a .dll, .so or .dylib. An app that adopts it owns the download, the cache and the unload.
Reality
The OpenAI bill it displaced was about EUR 10 a month. Paying that back takes a second set of calibration values, and the only comparison actually benchmarked in the writeup is embeddings.
Reality
A Llama 3.2 3B agent running offline invented a $1,990 balance on a $1,975 invoice. The open-sourced answer leaves prose to the model and hands every number to deterministic Python behind a tri-state router.
Reality
A 15M-parameter model streams English text on a 2007 PSP at about one token per second. That is the extreme end of a sizing rule. The harder half of that rule is checking whether the file that fits is a format its own maintainer recommends.
Reality
Stanford's AI Index puts a 142-fold parameter cut and a more than 280-fold price cut behind one fixed MMLU threshold. That narrows where building your own still pays, and it lands in a state-law count that doubled in a year.
Reality