wowsuch ai
Daily brief2026-09-24 UTC
Eikos-4B and Eikos-27B are open-weight models that return typed decisions and probabilities for finance and rule-based tasks; their sources state that they do not predict prices. [1] [2] Liquid AI introduced LFM2.5-VL-DSpark, an experimental vision-language draft model for speculative decoding with LFM2.5-VL-3B. [3] Other releases include a streaming Chinese and English speech model, structured-decision and local Apple Neural Engine models, and Viggle models for character replacement, camera views and image editing. [4] [5] [6] [7] [8] [9]
Anthropic added automatic evaluation of Managed Agents tool calls and evaluation metadata to the Claude Platform; its ant CLI also gained a beta command to connect a terminal to agent sessions. [11] JevHarness is a Python project and agent skill for building task-specific harnesses, optionally using rewards and execution traces, then running selected strategies without an authoring LLM on every decision. [10]
Runway introduced a DaVinci Resolve plugin that brings Runway tools into the video-editing application; it is available on paid plans. [12]
Each numbered link above points to one of these catalog records, with its sources.
Eikos-4B is an open-weight model that returns typed decisions with probability distributions and confidence values in a single forward pass. It is focused on finance, trading, and trade-finance rules, and does not predict prices.
Sources: Hugging Face, Hugging Face
Eikos-27B is an open typed-decision model that returns probabilities over bounded options in a single forward pass. It is trained for finance and rule-based decisions, and the source states that it does not predict prices.
Sources: Hugging Face, Hugging Face
LFM2.5-VL-DSpark is an experimental vision-language draft model for speculative decoding with LFM2.5-VL-3B. Liquid AI reports decoding throughput improvements of up to 2.66× on GPUs and 3.13× on edge devices, with output quality unchanged.
Sources: Liquid AI, Hugging Face
Audio8 ASR Infinite is a native streaming speech recognition model for Chinese and English. Its preview checkpoint supports selectable audio clocks, configurable transcription delay, and rolling KV caching for unlimited-length transcription.
Sources: Hugging Face
Bespoke-Nimble-9B is a Qwen3.5-9B LoRA adapter for context-grounded choices, booleans, and rubric score levels. It scores allowed answer tokens directly rather than generating reasoning or free-form answers.
Sources: Bespoke Labs
Viggle-Animate replaces a character in a video using the driving clip and one of its frames repainted with the desired character. The model propagates that edit through the clip without a pose estimator, segmenter, face tracker, or text encoder in the video stage.
Sources: Viggle
Meridian is a geometry-guided video model released as two LoRA adapters for the MiniMax-H3 transformer. It generates new camera viewpoints of existing video, and can also create camera movement from a single image.
Sources: Viggle
Qwen-Image-2.1-viggle-turbo v0.2.1 is a LoRA adapter distilled from Qwen-Image-2.1 for text-to-image generation and instruction-driven image editing. It supports editing with one to three reference images and is sampled in six transformer passes without classifier-free guidance.
Sources: Viggle
Laya Multilingual Core ML ANE is a downloadable FP16 checkpoint for short decisions, listed with CPU and Neural Engine support. The bundle has a B1/L96 capacity and a 96-token total limit.
Sources: GitHub new repositories
JevHarness is a Python project and agent skill for creating task-specific Jev harnesses with an authoring LLM. It can optionally use rewards and complete execution traces to improve a harness, then freeze the selected strategy so runtime decisions use the harness code and Jev calls without an authoring LLM on every decision.
Sources: GitHub new repositories
Anthropic updated the Claude Platform to let the server automatically evaluate Managed Agents tool calls via a new "auto" permission mode and to report evaluation metadata in agent events. The ant CLI (v1.32.0) gained a beta sessions connect command to attach a terminal to Managed Agents sessions and a --web option to open the session viewer locally.
Certain refusals that arrive before any output are now billed at the rates of the model that ran the request. The change applies on all platforms; other pre-output refusal categories remain unbilled, and fallback credit is unchanged.
The Runway DaVinci Resolve Plugin brings Runway tools into DaVinci Resolve. It is available on paid plans.
Sources: Runway
Written by openai/gpt-6-luna from the catalog records listed here; every sentence cites the records it rests on. Which records a brief covers is decided by fixed rules: the new developments and major releases of the window with the strongest evidence.
Original catalog content is licensed under CC BY 4.0, attributed to wowsuchai. Source quotations and linked content keep their original rights.