SShortSingh.
Back to feed

Anthropic's amended S-1 shows revenue surging 1,127% as compute costs erode operating margins

0
·2 views

Anthropic filed an amended S-1 with the U.S. SEC on Monday, disclosing third-quarter 2026 figures for the first time. Revenue reached $16.96 billion in the quarter, up 1,127% year over year, while computing and infrastructure costs rose to $9.86 billion. Operating income swung from a $974 million profit in the prior quarter to a $567 million loss, driven largely by short-term compute contracts priced above multi-year deals. Most of the quarter's large net loss of $23.29 billion came from non-cash fair value changes on convertible notes, derivative liabilities and SAFEs. The filing also reveals roughly $895 billion in binding purchase and lease commitments, with about $69.8 billion due within twelve months.

Read the full story at DEV Community

This is an AI-generated summary. ShortSingh links to the original source for the complete article.

Discussion (0)

Log in to join the discussion and vote.

Log in

Related stories

0
ProgrammingDEV Community ·

Developer swaps autonomous AI agent for deterministic routing to cut costs and boost control

A developer built an agent that autonomously managed a calendar scheduling assistant, but it repeatedly called the model and databases, making a single message cost over a dollar in tokens. Engineering flaws contributed, yet the developer judged the design unsustainable even after fixes. They rebuilt the system with the same user-facing behavior but a deterministic architecture of AI nodes and switches that tag workflow state and route messages. The new design reduced costs and gave the developer more control and better security by limiting the AI's autonomy, following the least-privilege principle. The author argues this principle deserves more attention from developers building AI systems.

0
ProgrammingDEV Community ·

Detecting Input Drift with PSI Before Model Accuracy Degrades

Deployed machine learning models can lose accuracy even when their code and weights stay unchanged, because the data they receive shifts over time. Ground-truth labels often arrive late, so monitoring input and output distributions offers an earlier warning sign. The article recommends using the Population Stability Index (PSI) to compare live data against a fixed, deliberately chosen baseline snapshot. It explains the PSI formula, offers a NumPy implementation, and suggests common threshold conventions that should be calibrated to each system. The author also advises sanity-checking the metric and measuring its natural noise on small samples before setting alerts.

0
ProgrammingDEV Community ·

Developer builds simple tool to flag salon leads left unanswered for 15 minutes

A developer built a small booking tool for a salon to address customers who leave enquiries that go unanswered while the owner is busy. The system has three parts: a booking form that saves each lead, a follow-up message preview generated immediately from the lead's details, and an owner inbox that shows leads newest first. If a lead goes 15 minutes without being marked as attended, the inbox displays a missed lead badge until the owner clicks to clear it. The backend uses Python with FastAPI and SQLite, while the front end is plain HTML and JavaScript without paid APIs. The developer found that the owner mainly needed a clear view of waiting leads rather than extra features, and the next planned step is connecting a real WhatsApp or SMS service to send the follow-up messages.

0
ProgrammingDEV Community ·

OnePassApps Stylesheet Analysis: 403 Custom Properties, Container Queries and Forced Colours

A team member at OnePassApps, which runs app.onepassapps.com, published a byte-level breakdown of the CSS its live site serves. The home page links a 3 KB pre-JavaScript fallback stylesheet and a 242 KB main stylesheet with long-lived caching. The main file defines 403 named custom properties grouped into families such as brand colours, ink scales, control sizes, motion timings and layout containers. Its 14 font-face blocks for Inter Variable are served from the same host as the CSS, with no third-party font requests and no font preloading. The author also notes that the stylesheet contains no prefers-color-scheme rules, so theme selection is handled in application JavaScript, and that it includes accessibility media queries for forced colours, contrast and reduced motion.