Search “Model releases”

Meta releases Muse Wonder and Muse Wonder Full, surpassing Claude on long-horizon coding tests…

Meta Superintelligence Labs launched its new-generation flagship models Muse Wonder and Muse Wonder Full on October 13. The Full version scored 56 on the Artificial Analysis Composite Intelligence Index, 2 points behind Claude Opus 5.5's 58. On the DeepSWE 1.1 long-horizon software engineering test, Wonder Full scored 79.0, surpassing Opus 5.5's 74.2 and GPT-6 Astra's 74.1, but it still trails Opus on Terminal-Bench 4.0 and ProgramBench. Artificial Analysis estimates its average task cost at $4.52, up 183% from the previous generation Muse Spark 1.3; a developer comparison on the same prompt shows its 3D game generation is noticeably worse than Opus, and Meta's model card has acknowledged 3D tasks as a known limitation.

Author trained six models in two days using ML Intern prompts

Hugging Face blogger yvrjsharma released seven prompts they used with ML Intern and showcased six projects completed within two days, including a citrus disease VLM based on Qwen3.5-2B, a Huggy character LoRA for FLUX.2 klein 4B, a camera viewpoint LoRA and a doodle LoRA for Qwen-Image 2.1, Pocket Rewriter 0.8B/2B distilled models, and a version of Agate Preview 002 distilled from 50 steps to 4 steps. The author recommends stating verified facts in prompts, requiring a pre-training baseline and smoke tests with checks, and setting spending caps, for example limiting the citrus project to 12 dollars. All prompts are available on GitHub at yvrjsharma/ml-intern-prompts, and ML Intern mode can be enabled in HuggingChat.

English translation is being prepared.

The Evolution of Qwen: From 7B to 2.4T, a Full Retrospective of Alibaba's AI Model Series

Alibaba Cloud's Qwen series began with Tongyi Qianwen in April 2023, open-sourced Qwen-7B in August 2023, launched the first Qwen3-Max-Preview with over 1 trillion parameters in September 2025, and ultimately evolved to an open-weight model with 2.4 trillion parameters. In 2024, three generations—Qwen1.5, Qwen2, and Qwen2.5—were released in succession; Qwen2.5's pre-training data grew from 7 trillion to 18 trillion tokens, and model downloads exceeded 40 million. In 2025, Qwen3 adopted a hybrid thinking mode, covering 119 languages, and Qwen3-Next-80B-A3B claims training costs of only 10% of Qwen3-32B and 10x higher inference throughput above 32K tokens.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

Google Releases Gemini 4 Argon, Topping Multiple Leaderboards and Focusing on Cybersecurity Defense

Google has released its new-generation flagship model Gemini 4 Argon, which scores 77.9% on DeepSWE v1.1, surpassing Claude Opus 5.5's 74.2% and GPT-6 Astra's 74.1%, and tops the Vals Index with 68.90%. The model supports output of up to one million tokens, and during the promotional period costs $2 per million input tokens and $10 per million output tokens, with a per-question cost of about 40% of Astra's. It is currently available only to vetted cybersecurity defense teams, with the first batch entering Google's Fairwind program.

OpenAI Team Responds to Computer Use Progress Doubts, Recaps One-Week Sprint on Decisions API

In a podcast after OpenAI DevDay, Ari Weinstein, who is in charge of the Computer Use agent, responded to Dwarkesh's doubts about Computer Use progressing slowly, saying it is now completely different from a few months ago because the model has become better at debugging and retrying, and can write JavaScript to execute multiple steps at once while combining screenshots, accessibility, DOM, and Playwright. He also mentioned that Dots provides each Dot with a cloud Linux virtual computer, and that Computer Use has entered the Agents API. Nikunj Handa, head of the OpenAI API team, introduced updates to GPT-6 such as asynchronous tool calling, mid-turn steering, WebSockets, UltraFast inference, prompt caching and pre-warming, and compaction, and said the Decisions API is based on Luna weights, with no new model trained, and took only about a week from the emergence of the idea to a prototype.

English translation is being prepared.

The Father of OpenAI Reasoning Noam Brown's Latest Interview: Mathematics Is Just an Appetizer for the Multi-Agent Era

OpenAI researcher and core o1 author Noam Brown said in the latest podcast that multi-agent systems are a way to change inference-time compute from serial scaling to parallel scaling, but in solving the Millennium Prize Problems, multi-agent systems account for at most 10% of the credit; the core is still that the model itself is sufficiently powerful. He revealed that when 5.6 was released, they had a truly formed multi-agent system for the first time, provided in the form of Ultra Mode, with 4 Agents by default and adjustable higher; with 4 Agents, the problem-solving speed is about twice as fast, and tests scaled up to as many as 16 Agents.

English translation is being prepared.

AIhub Monthly Roundup: September 2026 — Tracking Animal Populations, Recommender Systems, and an Interview with Ken Goldberg

AIhub has published its September 2026 monthly roundup, covering an interview with roboticist Ken Goldberg, tracking wildlife populations with deep learning, early-stage ranker optimization for large-scale recommender systems, and research on how external knowledge flows through AI systems. This month, the new GPT-series model GPT-6 Astra was made available to paying users, and Sebastian Raschka wrote about his first impressions; several Fields Medalists signed a statement warning that using generative AI for mathematical research involves a misalignment between purpose and results.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

Ant Ling open-sources two 6B models in the Ming-Image-0.1-Design series

Ant Ling has open-sourced the Ming-Image-0.1-Design series, comprising two 6B models: Design generates complete visual designs such as UIs, infographics, and posters from text requirements, while Layer decomposes a design image into 2–9 semantically independent RGBA layers; Design Skill and PPT Skill are also open-sourced. In the UI/UX Design evaluation updated by Artificial Analysis on September 18, 2026, Design ranks 1st among open-source models with an Elo score of 1082; Layer achieves the best results on all 12 metrics of Crello-Test, reducing single-inference time from 795 seconds to 183 seconds. After Ant's internal Huamei integrated Layer, the time to first obtain editable layers dropped from over 5 minutes to about 40 seconds.

Debating RSI, the China-US gap, and uneven model capabilities with Epoch AI's JS Denain

In episode 19 of the Interconnects podcast, Nathan Lambert and JS Denain, head of the insights team at Epoch AI, discuss RSI, the China-US model gap, and uneven capabilities. JS believes the evidence in public blog posts from OpenAI and Anthropic is insufficient to prove that AI is about to self-accelerate, but the 2x monthly growth in Codex research spending mentioned in OpenAI's blog post is worth noting; both hold highly uncertain views on where AI is heading.

TypeSafe CEO Diogo Almeida on Jev: Building a System One Model for Production, Not for God

InstructGPT paper author and TypeSafe CEO Diogo Almeida detailed his new model Jev on the Latent Space podcast, describing it as a System One large programmatic model oriented toward code consumption, with its core innovation being an unpublished RLCD (reinforcement learning for calibrated decision-making), optimized for "intelligence per dollar" rather than human preference feedback. He criticized RLHF for causing mode collapse, sycophancy, and hallucinations, arguing that safety alignment suits products like ChatGPT but fragments intelligence at the API layer, and said TypeSafe would rather be seen as a data lab, refusing to publish benchmark tests. He revealed that Jev's daily token consumption has exceeded one trillion since launch, and previewed follow-up models such as ReasoningJev.