Models & Releases Research Products & Apps Business & Funding

Modelwire

A curated feed of what matters in AI. Independent, ad-supported, built in Denver, Colorado.

Read

Today
Models & Releases
Research
Business & Funding

About

About Modelwire
Methodology
Our sources
Editor's notes
Contact
Advertise

Legal

Privacy policy
Terms of use
DMCA & takedowns
Corrections

© 2026 Modelwire. All article links go to the original publishers.Summaries generated by Modelwire. We don’t republish full articles.

Earlier stories

The full Modelwire feed, ordered by publish time.

Illustration for: DeepSeek-V4: a million-token context that agents can actually use

Models & Releases

DeepSeek-V4: a million-token context that agents can actually use

DeepSeek released V4 with a million-token context window, marking a significant expansion in how much information agents can process in a single session. The capability addresses a practical bottleneck for long-horizon reasoning and multi-step workflows.

Hugging Face·Apr 24

89

Illustration for: Anthropic and NEC collaborate to build Japan’s largest AI engineering workforce

Business & Funding

Anthropic and NEC collaborate to build Japan’s largest AI engineering workforce

Anthropic and NEC are partnering to develop AI engineering talent in Japan, positioning the collaboration as a path to building the country's largest AI workforce. The deal signals Anthropic's expansion into Asia-Pacific and reflects growing demand for localized AI expertise outside the US.

Anthropic·Apr 24

75

Illustration for: Claude is connecting directly to your personal apps like Spotify, Uber Eats, and TurboTax

Products & Apps

Claude is connecting directly to your personal apps like Spotify, Uber Eats, and TurboTax

Anthropic expanded Claude's app integrations beyond enterprise tools to personal services including Spotify, Uber Eats, TurboTax, Instacart, and AllTrails, letting users pull data directly into conversations. The move signals a shift toward consumer-facing AI utility and raises questions about data access and privacy at scale.

The Verge — AI·Apr 23

69

Illustration for: Extract PDF text in your browser with LiteParse for the web

Extract PDF text in your browser with LiteParse for the web

Simon Willison ported LlamaIndex's LiteParse PDF extraction tool to run entirely in the browser, preserving its non-AI approach to text parsing and OCR fallback. The browser version maintains compatibility with LiteParse's core libraries while enabling client-side PDF processing without server dependencies.

Simon Willison·Apr 23

64

Illustration for: Introducing GPT-5.5 with NVIDIA

Models & Releases Business & Funding

Introducing GPT-5.5 with NVIDIA

OpenAI unveiled GPT-5.5, positioning it as a task-execution model built in partnership with NVIDIA. The release highlights a new capability tier focused on autonomous problem-solving for engineering workflows, with early adoption from NVIDIA engineers.

OpenAI (YouTube)·Apr 23

97

Illustration for: First impressions of GPT-5.5 from Aaron Friel

Models & Releases

First impressions of GPT-5.5 from Aaron Friel

Aaron Friel from OpenAI discusses GPT-5.5's capabilities for handling faster, more autonomous long-running tasks in a conversation with Romain Huet. The interview explores how OpenAI teams are leveraging the new model's performance improvements for extended workflows.

OpenAI (YouTube)·Apr 23

89

Illustration for: Bret Taylor’s Sierra buys YC-backed AI startup Fragment

Business & Funding

Bret Taylor’s Sierra buys YC-backed AI startup Fragment

Sierra, Bret Taylor's AI customer service startup, has acquired Fragment, a YC-backed French AI firm. The deal signals consolidation in the competitive agent-software space as larger players absorb specialized capabilities.

TechCrunch — AI·Apr 23

58

Illustration for: First impressions of GPT-5.5 from Claire Vo

Models & Releases Products & Apps

First impressions of GPT-5.5 from Claire Vo

Claire Vo, founder of ChatPRD, shared early impressions of GPT-5.5 on OpenAI's channel, highlighting how the model enables new product workflows and helps resolve bugs in her own tool. The demo signals practical adoption patterns among AI-native builders.

OpenAI (YouTube)·Apr 23

77

Illustration for: First impressions of GPT-5.5 from Will Koh

Models & Releases Products & Apps

First impressions of GPT-5.5 from Will Koh

Will Koh from Ramp demonstrated how GPT-5.5 enhances the fintech platform's tool-use capabilities, enabling more intelligent function selection for customers. The upgrade signals meaningful progress in model reasoning for real-world B2B workflows.

OpenAI (YouTube)·Apr 23

77

Illustration for: Community Votes to Deny Water to Nuclear Weapons Data Center

Hardware & Infra Policy & Regulation

Community Votes to Deny Water to Nuclear Weapons Data Center

A township has imposed a one-year water moratorium on a new AI data center planned by U.S. nuclear weapons researchers, blocking infrastructure critical to large-scale compute operations before construction begins.

404 Media·Apr 23

69

Illustration for: A pelican for GPT-5.5 via the semi-official Codex backdoor API

Models & Releases Products & Apps

A pelican for GPT-5.5 via the semi-official Codex backdoor API

OpenAI released GPT-5.5 to Codex and paid ChatGPT users, with Simon Willison reporting strong performance on code generation tasks. The API remains unavailable pending safety review, delaying broader developer access.

Simon Willison·Apr 23

89

Illustration for: Meta is laying off 10 percent of its staff

Business & Funding

Meta is laying off 10 percent of its staff

Meta is cutting 10 percent of its workforce (roughly 8,000 employees) in May and freezing 6,000 open positions, per a memo from chief people officer Janelle Gale. The move follows the company's heavy spending on AI infrastructure and model development.

The Verge — AI·Apr 23

65

Illustration for: Meet Noscroll, an AI bot that does your doomscrolling for you

Products & Apps

Meet Noscroll, an AI bot that does your doomscrolling for you

Noscroll, a new AI bot, automates social media consumption by reading feeds on behalf of users to combat doomscrolling habits. The tool represents a niche application of AI to behavioral wellness rather than a technical or capability breakthrough.

TechCrunch — AI·Apr 23

47

Illustration for: llm-openai-via-codex 0.1a0

llm-openai-via-codex 0.1a0

Simon Willison released llm-openai-via-codex, a plugin that reuses Codex CLI credentials to route OpenAI API calls through the LLM command-line tool. The workaround lets developers access GPT models without separate authentication setup.

Simon Willison·Apr 23

64

Illustration for: OpenAI unveils GPT-5.5, claims a "new class of intelligence" at double the API price

Models & Releases Products & Apps

OpenAI unveils GPT-5.5, claims a "new class of intelligence" at double the API price

OpenAI released GPT-5.5, an agentic model capable of autonomously orchestrating multiple tools to solve complex tasks. The company doubled API pricing for the new model, positioning it as a step forward in autonomous reasoning capabilities.

The Decoder·Apr 23

92

Illustration for: OpenAI releases GPT-5.5, bringing company one step closer to an AI ‘superapp’

Models & Releases

OpenAI releases GPT-5.5, bringing company one step closer to an AI ‘superapp’

OpenAI unveiled GPT-5.5, claiming broad capability improvements across multiple domains. The release positions the company's push toward an integrated AI platform spanning multiple use cases and modalities.

TechCrunch — AI·Apr 23

81

Illustration for: Anthropic’s Mythos breach was humiliating

Models & Releases Business & Funding

Anthropic’s Mythos breach was humiliating

Anthropic's Claude Mythos model, which the company claimed was too dangerous for public release due to advanced cybersecurity capabilities, has leaked to unauthorized users. The breach undermines Anthropic's controlled rollout strategy and raises questions about the gap between internal safety claims and actual containment.

The Verge — AI·Apr 23

69

Illustration for: At 'AI Coachella,' Stanford Students Line Up to Learn From Silicon Valley Royalty

Opinion & Analysis

At 'AI Coachella,' Stanford Students Line Up to Learn From Silicon Valley Royalty

Stanford's CS 153 course has become a campus phenomenon, drawing students eager to learn directly from prominent Silicon Valley figures. The class has generated significant social media buzz, though it's sparked mixed reactions among the student body.

WIRED — AI·Apr 23

47

Illustration for: First impressions of GPT-5.5

Models & Releases

First impressions of GPT-5.5

OpenAI researchers Will Koh, Claire Voh, and Aaron Friel shared early reactions to GPT-5.5 in a YouTube video, offering insider perspective on the model's capabilities and performance relative to its predecessors.

OpenAI (YouTube)·Apr 23

89

Illustration for: Introducing GPT-5.5

Models & Releases Products & Apps

Introducing GPT-5.5

OpenAI unveiled GPT-5.5, positioning it as a new class of AI capable of handling complex multi-step tasks, tool use, and self-verification for agentic workflows. The model is now available in ChatGPT and Codex, signaling a shift toward AI systems that can autonomously complete real-world work.

OpenAI (YouTube)·Apr 23

100

Illustration for: GPU Renters Are Playing a Silicon Lottery

Hardware & Infra Research

GPU Renters Are Playing a Silicon Lottery

Research from William & Mary, Jefferson Lab, and Silicon Data reveals that identical GPU models exhibit substantial performance variance when rented from cloud providers, turning procurement into an unpredictable gamble for AI teams. The silicon lottery effect, where manufacturing tolerances create chip-to-chip differences, compounds cost uncertainty for organizations scaling compute infrastructure. This finding reshapes how practitioners should evaluate cloud GPU pricing and benchmarking claims, suggesting that nominal specs alone cannot guarantee consistent training or inference economics.

IEEE Spectrum - AI·Apr 23

69

Illustration for: AWS Bets on Frontier Agents as the Next Era of Enterprise AI

Business & Funding Products & Apps

AWS Bets on Frontier Agents as the Next Era of Enterprise AI

AWS is positioning autonomous, long-running agents as enterprise AI's next inflection point, signaling a shift from single-task models to persistent, goal-oriented systems that operate across workflows.

AI Business·Apr 23

61

Illustration for: Trump science advisor says Chinese actors are copying American AI at massive scale

Policy & Regulation

Trump science advisor says Chinese actors are copying American AI at massive scale

The Trump administration claims evidence of systematic model distillation attacks by Chinese actors targeting US frontier AI systems at industrial scale, signaling a shift toward more aggressive AI supply-chain defense.

The Decoder·Apr 23

73

Illustration for: OpenAI says its new GPT-5.5 model is more efficient and better at coding

Models & Releases

OpenAI says its new GPT-5.5 model is more efficient and better at coding

OpenAI released GPT-5.5, claiming improvements in efficiency and code generation over last month's GPT-5.4. The rapid iteration signals intensifying competition in frontier model capability, though concrete benchmarks remain absent from the announcement.

The Verge — AI·Apr 23

69

Illustration for: Temporal Taskification in Streaming Continual Learning: A Source of Evaluation Instability

Temporal Taskification in Streaming Continual Learning: A Source of Evaluation Instability

Researchers show that how continuous data streams are split into tasks during continual learning evaluation significantly alters benchmark results, introducing hidden instability independent of model choice. They propose metrics to diagnose taskification sensitivity and test the effect across four major CL algorithms.

arXiv cs.LG·Apr 23

58

Illustration for: Evaluation of Automatic Speech Recognition Using Generative Large Language Models

Evaluation of Automatic Speech Recognition Using Generative Large Language Models

Researchers show decoder-based LLMs can evaluate speech recognition quality far better than traditional metrics, achieving 92-94% agreement with human judges on the HATS dataset versus 63% for Word Error Rate. The finding suggests generative models offer a practical alternative to semantic embeddings for ASR evaluation.

arXiv cs.CL·Apr 23

58

Illustration for: Fine-Tuning Regimes Define Distinct Continual Learning Problems

Fine-Tuning Regimes Define Distinct Continual Learning Problems

Researchers show that how models are fine-tuned during continual learning fundamentally changes the problem itself, not just the solution. By varying which parameters remain trainable across sequential tasks, the effective learning dynamics shift, suggesting current benchmarks may unfairly compare methods across incompatible regimes.

arXiv cs.LG·Apr 23

52

Illustration for: The Sample Complexity of Multicalibration

The Sample Complexity of Multicalibration

Researchers prove that achieving multicalibration in batch learning requires Θ(ε⁻³) samples, establishing a fundamental separation from marginal calibration's lower complexity. The result combines information-theoretic lower bounds with a practical randomized algorithm via online-to-batch conversion.

arXiv cs.LG·Apr 23

52

Illustration for: MathDuels: Evaluating LLMs as Problem Posers and Solvers

Research Models & Releases

MathDuels: Evaluating LLMs as Problem Posers and Solvers

Researchers introduce MathDuels, a self-play benchmark where LLMs both author and solve math problems to differentiate model capabilities beyond static benchmarks. The framework uses adversarial prompting and a Rasch model to jointly measure solver ability and problem difficulty, addressing ceiling effects in existing evaluations.

arXiv cs.CL·Apr 23

62

Illustration for: When Prompts Override Vision: Prompt-Induced Hallucinations in LVLMs

When Prompts Override Vision: Prompt-Induced Hallucinations in LVLMs

Researchers introduce HalluScope, a benchmark revealing that vision-language model hallucinations stem primarily from textual priors baked into prompts rather than vision backbone limitations. The finding reframes how teams should approach reducing false outputs in multimodal systems.

arXiv cs.CL·Apr 23

62

Older stories →