BottleCap AI has released ThinkingCap-Qwen3.8-27B, a fine-tune of Qwen3.8-27B that spends 37.2% fewer thinking tokens across 12 benchmarks. Macro accuracy moves from 86.65% to 85.79%, and long-context AA-LCR improves by 2.25pp. The model is a drop-in replacement on vLLM and SGLang, with FP8, NVFP4, GGUF and MLX builds.

The post BottleCap AI Releases ThinkingCap-Qwen3.8-27B: 37.2% Fewer Thinking Tokens at a 0.86pp Accuracy Cost appeared first on MarkTechPost.

Contrastive-LM has released CLM-8B, an open System One model that scores candidate actions against a state instead of generating text. It adds 2 small projection heads to a frozen Qwen3-8B encoder and trains them with a contrastive InfoNCE objective. In zero-shot tests it runs up to 9× faster than TypeSafe's Jev. With fine-tuned heads as a verifier, it reaches 81.6% on held-out DeepSWE tasks and 87.6% on held-out Terminal-Bench 2.1 tasks.

The post Contrastive-LM Releases CLM-8B: An Open System One Model That Scores Agent Actions Up to 9× Faster Than Jev appeared first on MarkTechPost.

This tutorial provides a complete coding guide to TypeSafe AI's Jev, a System One model designed for non-text, structured judgments. It covers installing the official Python SDK, using primitive question types (Choice, Score, Noul), implementing speculative fan-out, confidence-gated routing, and building async production workflows

The post A Coding Guide to TypeSafe AI Jev: Typed Decisions, Calibrated Confidence, and Speculative Fan-Out with a System One Model appeared first on MarkTechPost.

Google has released Gemini 3.8 Flash TTS and Flash-Lite TTS, 2 new text-to-speech models available now through the Gemini API and Google AI Studio. Flash TTS designs new voices from natural language prompts across 100+ languages. It ranks #1 on Hume AI's Voice Design Benchmark with a score of 71.4. Flash-Lite TTS targets high-volume dubbing and voice agents at lower cost.

The post Google Releases Gemini 3.8 Flash TTS and Flash-Lite TTS With Prompt-Based Voice Design appeared first on MarkTechPost.

NVIDIA has released Nemotron 3 Diarization, an open-weight speaker diarization model on Hugging Face. It answers one question about any conversation: who spoke when. The 100M-parameter model tracks up to 8 speakers, including when voices overlap. One checkpoint handles both offline recordings and real-time streaming. Is it deployable? Yes. The weights are released under the […]

The post NVIDIA Releases Nemotron 3 Diarization: A 100M-Parameter Open-Weight Model That Tracks 8 Speakers in Real Time appeared first on MarkTechPost.

Nokia’s applied research team has open-sourced AnyJev, a Python library that turns an open LLM into a decision model. It needs no training. It targets a common production job: picking one answer from a fixed set instead of writing a sentence. Is it deployable? Yes, it installs from PyPI, ships under Apache-2.0, and has transformers […]

The post Nokia Open-Sources AnyJev: A Training-Free Layer That Turns Any Open LLM Into a Calibrated Decision Model appeared first on MarkTechPost.

Kyutai has released Voice of Reason, 2 open-weight speech-to-speech models built on GLM-4-Voice-9B. Supervised fine-tuning and reinforcement learning lift spoken GSM8K accuracy from 27.3% to 77.1%. There is no transcription step and no text LLM in the loop. Both checkpoints are on Hugging Face and run on a single H100.

The post Kyutai Releases Voice of Reason: A Speech-Native Model that Solves Spoken Math with Reinforcement Learning appeared first on MarkTechPost.

OpenAI has released GPT-6 Sol and GPT-6 Luna, 2 lower-cost models trained with methods similar to GPT-6 Astra. Sol costs $2/$10 and Luna $0.10/$0.50 per 1M tokens. Both are available now in the API, ChatGPT Work and Codex. They come with improved prompt caching for long-running agents.

The post OpenAI Releases GPT-6 Sol and Luna: 50% Cheaper API Pricing and Benchmarks appeared first on MarkTechPost.

Voice input on phones has been solved for years. What has not been solved is the output. Speak into most dictation tools and you get back exactly what you said, fillers and false starts included, in a note you then have to clean up and move somewhere else. SpeakON attacks that gap with hardware: a […]

The post SpeakON Ships a MagSafe AI Voice Button With Its Own Microphone appeared first on MarkTechPost.

Anthropic has released Claude Opus 5.5, the first model in its new Claude 5.5 family. The team states it performs at the level of Claude Fable 5.1 on most work. It also costs 40% less to run than Opus 5 on typical workloads at default settings. On Anthropic’s own benchmarks, it leads in agentic coding, […]

The post Anthropic Releases Claude Opus 5.5: Fable 5.1-Level Performance at 40% Lower Running Cost Than Opus 5 appeared first on MarkTechPost.

Eligible iPhone owners in the United States can begin submitting claims on September 21, 2026, in the $250 million settlement of Landsheft v. Apple Inc., a class action over Siri Apple Intelligence features, according to the case's official settlement website. The settlement covers U.S. residents who were the original purchasers of an iPhone 15 Pro, iPhone 15 Pro Max, or any iPhone 16 model (iPhone 16, 16e, 16 Plus, 16 Pro, or 16 Pro Max) bought in the United States between June 10, 2024, and…

BigID on September 21, 2026 announced the launch of AgentIQ, an agentic automation interface that lets customers operate their data security and compliance programs from a prompt or an AI agent, either inside BigID or directly from Claude, Copilot, GPT, or Gemini. What AgentIQ Does The release describes AgentIQ as a fully agentic automation interface to BigID. Customers can write custom prompts or use pre-built BigID agents to retrieve, report, respond, or remediate data and AI security and…

Simbe Robotics announced on September 21, 2026 that it has surpassed 3,000 autonomous units under contract, which the company described as the largest commercially committed fleet of shelf intelligence technology identified in company research. The announcement comes in the decade since Simbe introduced Tally, which the company calls the world's first inventory robot, and the company said it is building on that position to evolve from autonomous shelf digitization toward a broader platform for…

Google opened pre-orders for Googlebook on September 21, 2026, bringing its new laptop line to market with five flagship models built by Acer, ASUS, Dell, HP, and Lenovo at a starting price of $899. The announcement came in a post by John Solomon, Google's vice president and general manager for Laptops, Tablets & Android Enterprise. Googlebook is built on the Android technology stack paired with desktop foundations from ChromeOS, and it is designed to work with an Android phone from the first…

The Independent International Scientific Panel on AI published a thematic brief on September 21, 2026, that describes the May–July 2026 OpenAI-Hugging Face incident as an early warning of one possible route to more severe future loss of human control over artificial intelligence: capable agents persistently pursuing goals that conflict with human intentions. The brief, AI Agents, Misalignment and the Risk of Losing Human Control: Evidence from the OpenAI-Hugging Face Incident, states that…

A loss function converts the difference between predictions and targets into a quantity that learning algorithms try to minimize. This guide explains the mechanism, trade-offs, evaluation, and controls that matter in practice.

I've spent most of my career around people who work with their hands, including the electricians and warehouse workers who show up at 5 a.m. and don't leave until the job's done. Frankly, I'll tell you what I hear from almost every one of them when the subject of AI comes up: some version of “great, another thing that's going to replace me”. I understand why they think that. Every headline about artificial intelligence sounds like it was written by someone who's never worked a shift that…

Any intervention in a bank's core system makes an engineering team wince. Legacy solutions are rigid, so a single change can take a month to move through. Lending, which is one of a bank's major sources of income, feels this most. Anyone who has shipped a release on a bank core knows the pattern: a new product needs a new approval tier, a revised document checklist, an adjusted scoring rule, and an updated reporting feed, each touching code that has run for decades. As a result, new products…

Retail media optimization platform Quartile announced on September 21, 2026 that it has become a technology partner supporting advertising in ChatGPT, giving participating advertisers access to ChatGPT Ads through Quartile's existing tools and workflows. In the release, datelined New York, Quartile, which describes itself as the world's largest retail media optimization platform, said the partnership expands its AI-powered advertising capabilities into conversational discovery. The company said…

We cloned one 10-second voice on 7 platforms and ranked them on reference audio, consent, licensing, and cost.

The post Best Voice Cloning APIs in 2026: Speaker Similarity, Consent Checks, and Price per 1M Characters appeared first on MarkTechPost.