Latest news, beside the leaderboard

On Oct 28 independent evaluator Artificial Analysis released Intelligence Index v5, its biggest revision since v4 in January: it adds Cognition's FrontierCode 2.0 and Terminal-Bench-Science, swaps Humanity's Last Exam for HLE-Diamond, and now draws 60% of its weight from private or held-out test sets. Claude Opus 5.5 (max) leads at 58, ahead of GPT-6 Astra at 57 and GPT-6.1 Sol at 55; the top open-weights model, Kimi K3.1 Pro (weights not yet released), scores 46, widening the gap from 3 points in v4.3 to 12. It also launched Progress Index v1 for comparing models across years.

On Oct 28 Meta announced that Muse, its personal AI agent, is now available in Singapore on iOS, Android and muse.ai. Beyond Gmail, Google Calendar and Instagram, the first local connector partners are Shopee, Grab, Singapore Airlines, Agoda, ShopBack, Klook, Trip.com and Lazada, spanning shopping, groceries, dining and travel, with some connector features rolling out after launch. Muse runs on Muse Secure VM, a dedicated cloud computer that keeps passwords and payment methods out of the agent's view, and it asks before sensitive actions such as sending an email or making a purchase. It is free for most uses, with paid subscription plans.

On Oct 27 Anthropic's Alignment team published a randomized controlled trial, "How AI agents shape experts' judgment calls", shared on X by researcher Alex Tamkin. Fifty analysts experienced in causal inference used Claude Code (Claude Opus 4.8) to evaluate a program for a fictional farming cooperative, and on each analytical decision the agent's lean was set by coin flip without the analyst being told. Their choices matched the agent's lean 86% of the time (about 50% would mean no effect), and 77% later repeated the agent's arguments while only 15% credited the agent.

On Oct 27, Greg Isenberg (@gregisenberg, about 752K followers), founder and CEO of Late Checkout, posted on X: "saas is becoming free: give away the crm, give away the dashboard, give away the app, and charge for the agent that does the work." Within about seven hours the post had about 109K views, 1,526 likes, 277 replies and 37 quotes, with many founders debating whether per-seat pricing can really give way to pricing for the work an agent completes. This is a discussion of an opinion, not a pricing announcement from any company.

On Oct 27 Ant Ling (inclusionAI) open-sourced Ling-3.1-flash-1019, a further post-trained upgrade of the early-October Ling-3.1-flash: 560B total parameters, 25B active per token, a 1M-token context and an MIT license. Ant says it scores 44 on the Artificial Analysis Intelligence Index v4.3.2 (up from 41), with clear gains on agent tasks such as AA-Briefcase and AutomationBench-AA. It also released Ling-3.1-flash-base, an open foundation checkpoint for further training.

In an Oct 27 blog post, Anthropic described months of deep research partnerships between its life sciences team and six organizations (BigHat, Bristol Myers Squibb, Divergence Labs, Enveda, Generate Biomedicines and The Column Group), covering hit generation, lead optimization, manufacturing paperwork and clinical trial planning. It also said it acquired the one-person company Mithril Medicine to run its own programs for rare childhood muscular dystrophy. Anthropic stresses the work is mostly preclinical so far.

On Oct 27 Cursor announced Annotate for Design Mode: in the Agents Window browser you can click elements, box an area, draw markup, edit text in place or tweak CSS, and each note is sent to the agent with element details and a screenshot so it can edit the code. The post says you can also review designs at different screen sizes. It needs Cursor 3.24.18 or later, and the shortcut is Cmd+Shift+D.

On Oct 27 Anthropic announced a new OneGov agreement with the US General Services Administration (GSA): government customers get 40% off Claude Enterprise, Claude for Government and the Claude API. For the first time the discount covers every product surface (direct purchases today, cloud-provider channels in the coming months) and extends to state, local, tribal and territorial governments. Anthropic estimates about $1 billion in savings, with no minimum spend and pay-as-you-go usage.

On Oct 27 Liquid AI released LFM3, a new generation of foundation models built on a tri-hybrid architecture of gated short convolutions, sliding-window attention and global attention, pre-trained on 22T tokens. The first six models all branch from LFM3-1.7B-Base at around 2B parameters, with weights on Hugging Face under the LFM Open License v1.0. Liquid says the text model LFM3-1.7B needs about 1 GB at 4-bit, supports 256K context and beats models several times its size on several agent benchmarks.

Cognition, the maker of Devin, released FrontierCode 2.0 on Oct 27. It keeps the method of maintainer-written tasks graded on whether the code is actually mergeable, but draws tasks from more repositories, adds security patches and performance work, uses deliberately underspecified prompts and adds many thoroughness criteria. In Cognition's results Claude Opus 5.5 leads at 45.5%, GPT-6 Astra follows at 42.8%, and Grok 4.7 and Gemini 4 Argon drop into the 20s, leaving the benchmark far from saturated.

On Oct 27 OpenBMB (ModelBest) open-sourced the MiniCPM-OCR suite with two models: MiniCPM-OCR 1.0, a 1.3B-parameter document parser that scores 93.68 on Real5-OmniDocBench and 54.7 on OCRBench v2 (vendor-reported), and MiniCPM-OCR Jev, a 2B OCR quality judge that scores any parser's output region by region from 0 to 1 and flags content that needs human review. Code and weights are Apache 2.0, with day-one support for vLLM, SGLang, llama.cpp, Ollama and multiple chip platforms.

Google DeepMind announced Genie 4 on Oct 27, calling it its most realistic and interactive world model yet. Built on Gemini Omni, it improves physics and human motion, natively generates ambient audio and interactive character dialogue for the first time, and uses Gemini 3.8 Flash in the loop to plan complex events. It is a research preview for trusted testers for now, coming soon to Google AI Ultra (Project Genie) and early access partners in the Gemini API and AI Studio; no quantitative benchmarks were published.

Microsoft AI announced on Oct 27 that MAI-Image-2.6 and MAI-Image-2.6-Flash are generally available in Microsoft Foundry, with multi-reference editing from up to five images, Bing web grounding, output up to 1.5K and published per-token pricing. The same day its next image editing model, MAI-Image-2.7-Preview, reached #3 overall in Arena's Image Edit leaderboard at 1470 points, 45 points above 2.6; for now it is available only for editing in MAI Playground.

AI math company Axiom announced on Oct 27 that swarms of dozens of AI agents formalized the Classification of Finite Simple Groups (CFSG; the original proof runs about 15,000 pages across 500+ articles by about 100 mathematicians) in the Lean proof assistant in 24 days, producing about 43.1M lines of Lean and 2.05M declarations, with the repo public on GitHub. The work surfaced 34 noteworthy deviations from the published record, and AxiomProver produced a new alternative argument for the Unbalanced Group Theorem.

On October 27 Apple announced a major Apple Creator Studio update: Logic Pro for Mac plus Keynote, Pages and Numbers gain beta intelligent assistants that Apple says are powered by a combination of Apple-built models and OpenAI and understand document context. Keynote, Pages and Numbers also add Model Context Protocol (MCP) support, letting external AI agents read content, make edits and run automated actions inside the apps. Pixelmator Pro adds natural-language Generate Fill, and Keynote introduces the open JSON-based .keyx format. The update ships free to existing subscribers on the App Store in November; pricing stays $12.99 per month or $129 per year.

Shanghai AI Laboratory has moved Atria Dawn from Preview to a full release, with open weights (plus an FP8 build) on Hugging Face under internlm/Atria-Dawn and on ModelScope, MIT licensed. Built on the 744B-parameter MoE GLM-5.2, it targets long-horizon research and engineering agent work; context doubles from the Preview's 256K to 512K and input is text only. The vendor table shows it leading the listed models on AutomationBench, CyberGym and SkillsBench, while still trailing Claude Opus 5 on coding tests such as SWE-bench Pro and Terminal-Bench 2.1 (all vendor-reported).

Anthropic published the paper "A 2^o(n) Time Algorithm for 3-SAT" (credited to Claude, Anthropic) and an AI-written Lean 4 formalization in its official anthropics/formal-math GitHub repo. It claims a deterministic algorithm solves 3-SAT in 2^o(n) time, which would refute the Exponential Time Hypothesis (ETH). Anthropic's Levent Alpoge said on X that an internal Claude research model found the core algorithm on Oct 16 with 14M output tokens and no human input. The repo stresses the result is purely asymptotic and says nothing about solving formulas in practice.

On Oct 27, Baidu's official @Baidu_Inc account announced the open-source release of Unlimited OCR 2 on Hugging Face under the MIT license. Its new SR-SWA attention compresses text-visual interactions into a fixed-size reference set, which Baidu says lets it parse documents of any length within a finite context window. Baidu reports 84.6 on MPDocBench (dozens of pages) and 77.1 on MarathonDocBench (hundreds of pages), both self-reported.

In an Oct 26 incident report on its blog, Cloudflare said that between Aug 27 and Sep 9, 155 responses from GLM-5.3 on Workers AI (@cf/zai-org/glm-5.3) to prompts longer than 681,280 tokens contained data from other customers' recent requests, all returned with HTTP 200. The cause was a prefill KV cache pool smaller than the 1,048,576-token context window, combined with an SGLang bug in disaggregated prefill/decode mode. Cloudflare found no evidence of deliberate exploitation, has fixed it, and recommends operators upgrade to SGLang v0.5.22 or later.

Cursor announced on Oct 26 (UTC) that DeepSeek V4.1 Flash is now available in Cursor and can be turned on under Cursor Settings > Models. Cursor says every effort level sits on the CursorBench cost frontier, from $0.60 to $0.98 per task, with no model scoring higher at those prices. The official CursorBench 4.0 leaderboard lists DeepSeek V4.1 Flash 1017 at 47.7% on Max ($0.98 per task), 46.3% on Medium ($0.80), 45.6% on High ($0.77) and 44.9% on Low ($0.60). The top score on the board, Opus 5.5 Max, is 57.8% but costs $13.43 per task.