Search “Model releases”

English translation is being prepared.

Anthropic's Engineering and Product Leads Discuss the Opus 5.5 Release and Claude's Agent Architecture

In a Latent Space podcast interview, Anthropic engineering lead Katelyn Lesse and product lead Angela Jiang discussed the Opus 5.5 release, covering adjustments to the model's writing style, pricing based on intelligence per dollar, and API changes such as requiring thinking blocks to be retained across turns to prevent distillation. The interview also described Claude Managed Agents' approach of separating sessions and the harness from the sandbox that executes code, with credentials mounted separately through a proxy so that a sandbox failure does not take down the entire session; Dreaming, meanwhile, is implemented as one managed agent inspecting another agent's history and writing back to memory. The team said that as safety work for model releases has grown heavier, recent efforts have shifted more toward safety, with progress on features such as skills and memory slowing relatively.

English translation is being prepared.

Ethan Mollick reviews 30 months of AI development in charts

On the release day of his new book Co-Existence, Ethan Mollick used charts to review the AI progress in the 30 months since his previous book Co-Intelligence was published. He said that since GPT-4, 20 models have successively taken the top spot, and 180 models were stronger than GPT-4 at release, 19 of which can run on modern smartphones; the METR long-task benchmark showed that the amount of human work AI can complete increased from 4 minutes in 2024 to 17 hours in April 2026, and the benchmark was subsequently discontinued because it could no longer measure accurately. He proposed a third mode of collaboration beyond centaur and cyborg: conjurer, that is, summoning agents, assigning tasks, and evaluating results. The charts in this article were completed by Opus 5.5 in Claude Projects orchestrating multiple agents.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

English translation is being prepared.

Kimi K3.1 Pro hands-on test: Why does quota consumption still spark controversy after the API price cut?

Kimi released K3.1 Pro on October 24, cutting the API input price from 20 yuan to 13 yuan per million tokens and the output price from 100 yuan to 68 yuan, but users questioned the cost-effectiveness due to rapid token consumption; on October 26, the company apologized, compensated quotas, and adjusted the model configuration. The author used a Pro subscription account to test five creative tasks, which consumed a total of about 25.48% of the monthly quota, with "The Peach Blossom Spring" at 6.91% the highest and the mechanical keyboard at 1.46% the lowest; consumption mainly came from Plugin entries and multiple rounds of revisions. The article also mentions the controversy over overseas bloggers being asked to hide paid partnership labels in promotions, and that former miHoYo international president Jin Wenyi joined Moonshot AI to oversee overseas expansion.

Anthropic S-1 shows roughly $850 billion in compute commitments over the next two decades

The Verge analyzed Anthropic's S-1 filing, noting that its non-cancelable purchase commitments, not-yet-commenced lease agreements, and new compute agreements total roughly $850 billion, and would approach $1 trillion if un finalized deals are included. The filing shows that revenue in the first half of 2026 grew 14-fold year over year to $16.3 billion, adjusted revenue was $1.62 billion, and net revenue retention was 900%. The author argues that these fixed costs require Anthropic to maintain rapid growth, and that the recursive self-improvement mentioned among its risk factors is precisely the key supporting this growth assumption.

OpenAI Developer Lead Discusses How to Choose Between Sol, Astra, and dots After DevDay

Vaibhav Srivastav, OpenAI's Head of Developer Experience and Community, said in an interview that GPT-6.1 Sol can achieve about 95% of the performance of GPT-6 Astra at only one-fifth the cost, and that he himself mainly uses Sol in his daily work. He recommends using Astra for adversarial review on new projects or unfamiliar problems, and said the Ultrafast tier targets real-time scenarios such as fraud detection and financial investigations, with speeds exceeding 300 tokens per second. He also described how the always-on agent OpenAI dots aggregates information from Slack, GitHub, Gmail, and calendars and handles PRs on users' behalf, as well as how Auto-review checks model actions against custom policies before they are taken.

Podcast: Kimi K3.1 Pro, the new cybersecurity reality, and the Muse user experience

In episode 20 of the Interconnects podcast, Nathan Lambert and Florian Brand discussed how Kimi K3.1 Pro ranks high on Artificial Analysis but receives mixed reviews, arguing that its training leans more toward exploitation than exploration, with weaker performance on long-horizon benchmarks such as ProgramBench. The two also discussed the cybersecurity controversy surrounding open-weight models, noting that no large-scale exploitation occurred after the disclosure of the PHP2Shell vulnerability, and talked about Anthropic not hiring junior employees as well as Florian's experience using Muse to handle everyday tasks such as shopping.

Behind the Kimi K3.1 Pro quota controversy: the stronger the model, the tighter the compute

Moonshot AI released Kimi K3.1 Pro on October 24, scoring 65.7% on Terminal-Bench 4.0, but users of the 199-yuan and 699-yuan plans complain that quotas are insufficient, and some developers estimate that the API value of the old plans is about 1.94 times that of the new plans. K3.1 Pro's per-million-Token input and output prices dropped from K3's 20 yuan and 100 yuan to 13 yuan and 68 yuan, while complex Agent tasks can consume millions of Tokens. A CAICT report shows that in June 2026, domestic daily average Tokens were about 500 trillion, nearly four times the figure in December 2025; the author believes that the compute shortage will still constrain independent model companies over the long term.

Databricks on AI coding cost governance: unit costs cut by up to 90%

Patrick, a product lead at Databricks, reviewed the cost overrun problem after the company fully adopted AI coding agents, saying that by migrating to cheaper models, routing, spend visibility, and cutting token overhead, unit costs in some scenarios fell by as much as 90%. He noted that hard budget caps would interrupt the best developers, so the company switched to gradual friction through warnings, approvals, and model downgrades, and productized its internal solution as Unity Gateway. The team built its own evaluation using real PRs from its own codebase, finding that GLM 5.2 matched Opus 4.8 in quality at about one-third lower cost per task, while a model that was 1.7 times cheaper per token ended up costing more per task because it consumed more tokens.

Anthropic CEO Dario Responds to Safety Concerns in a Long Post, Reveals High-End Model Was Held Back for Four Months

Anthropic CEO Dario posted a long message on X responding to questions from AI safety investigator Nathan Calvin, disclosing for the first time that the company had already trained a high-end model at the Mythos level as early as February this year. Because its cybersecurity defense mechanisms were not yet complete, it was released only to a very small number of partners, and was not opened to the public until the safety guardrails were in place in June—about four months in total. He also listed the costs the company has proactively borne for safety, including a large number of false positives caused by biological risk filtering, pushing for legislation such as California SB 53, and a plan to commit 5% of the company's shares to a public interest fund. Dario said this approach amounts to competing toward a higher standard, while Calvin questioned how much more the company can concede in the race for the sake of safety.

Nvidia Joins Six Major Financial Institutions to Build a $500 Billion+ AI Financing Platform

In August, Nvidia teamed up with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to establish a financing platform that can mobilize more than $500 billion for AI infrastructure companies. The author, Mehran Gul, argues that the key variable in the AI race is shifting from models and chips to financing capability; the five major US cloud providers are expected to spend $730 billion in capital expenditure in 2026, with up to one-third of that funded by debt issuance. The article also notes that Europe has about €10 trillion in household savings but lacks channels to convert savings into investment, while China faces the problem of computing power sitting idle after it is built.

English translation is being prepared.