#cost-tracking
-
AI News Briefing — Stripe confirms the OpenRouter deal, price still unofficial
Stripe confirmed the OpenRouter acquisition three days after Bloomberg broke it, without naming a price; OpenRouter says its commitments hold and it runs independently. OpenAI previews zero-retention abuse detection for frontier models.
-
AI News Briefing — August 18, 2026
GitHub broke for three hours and nineteen minutes; Cursor shipped Origin, its own git forge, the same afternoon — on by default for paid plans. Gartner puts agentic inference cost up fivefold by 2028.
-
AI News Briefing — August 17, 2026
Composio ran DeepSeek's leaderboard-topping V4 Flash through eight agent harnesses and got 53.8% task completion — only six of thirty workflows finished everywhere. AWS open-sourced a Cedar dialect that reasons about an agent's past tool calls.
-
AI News Briefing — August 16, 2026
SpaceX closed its $60 billion acquisition of Cursor, whose agents had passed an audited agent-security standard the day before. Anthropic's text watermarks, meanwhile, will barely mark generated code.
-
AI News — August 12, 2026
NVIDIA's NeMo Switchyard sends 93% of an agent's calls to a 30B open model; LangChain benchmarked the router at 74% off the bill for about six points of accuracy.
-
AI News — August 8, 2026
OpenAI says it cannot rule out that its next model, Astra, has critical cyber capabilities — the first time its Preparedness Framework has reached that level — and has paused internal work that lacks safeguards.
-
AI News — August 6, 2026
Meta shipped its first coding agent: Muse Code runs in the terminal on the new Muse Spark 1.2, keeps background agents alive across a whole session, and undercuts on price with a feedback-for-discount tier.
-
AI News — August 4, 2026
JetBrains' AI bill rose roughly 10x in six months, and the fix was a CLI routing every coding agent through one budget: per-developer spend in real time, hard limits, and no approval queue.
-
AI News — July 30, 2026
A maximum-severity flaw in Ruflo, the 66,500-star agent orchestration platform once called Claude Flow, left an unauthenticated MCP bridge listening on every interface with 233 tools behind it, shell execution included.
-
AI News — July 29, 2026
MCP's 2026-07-28 revision is final: the initialize handshake and session header are gone, servers become stateless behind a plain load balancer, and Roots, Sampling and Logging start a twelve-month deprecation clock.
-
AI News — July 27, 2026
Kimi K3's weights are not out yet: Moonshot's own Hugging Face repo carries a release timer set to 15:00 UTC today, hours after outlets began treating the drop as done.
-
AI News — July 8, 2026
US companies are routing a growing share of their AI traffic to cheaper Chinese open-weight models as frontier prices climb: CNBC reports the Chinese-model share of US tokens on OpenRouter has held above 30% every week since February and spiked as high as 46%, led by Z.ai's GLM 5.2 — within a point of Opus 4.8 on one agentic benchmark at roughly a fifth the cost — even as OpenAI shipped gpt-realtime-2.1 to add reasoning to voice agents and cut Realtime latency 25%.
-
AI News — July 7, 2026
The number behind Fable 5's return finally landed: with the 50%-of-weekly-limits promo expiring today, Anthropic's most powerful public model reverts to metered usage credits at its full $10/$50-per-million-token API rate — double Opus 4.8 and the steepest Anthropic has ever listed for a generally available model — turning the model that anchored a month of export-control drama into a per-call budget line the moment the promo window closed.
-
AI News — July 6, 2026
OpenAI's workspace agents left free preview today: the extended free window ended July 6 and credit-based metering is now live for agent runs invoked inside ChatGPT — a typical GPT-5.5 run burns 5–25 credits — turning the always-on Codex-powered agents into a budgeted per-run line item, while Claude Code's v2.1.200–201 renamed its permission "default" mode to "Manual" and landed another batch of background-agent reliability fixes.
-
AI News — July 4, 2026
A quiet US-holiday window, and the substantive item is governance, not a model: Anthropic shipped admin analytics, model-level entitlements, and spend-threshold alerts for Claude Enterprise, with usage-and-cost data exposed through an Analytics API that pipes into Datadog Cloud Cost Management and CloudZero — turning per-user, per-model Claude spend into something finance can see next to the rest of its cloud bill.
-
AI News — June 28, 2026
A day after the government cleared Mythos 5 for critical-infrastructure defenders, Axios reported the administration is close to letting Anthropic restore Fable 5 too — possibly as soon as this coming week, with talks continuing over the weekend and only the Pentagon and NSA still to sign off — even as CNBC documented enterprises pulling back from "tokenmaxxing": Lindy moved 100% of its traffic off Claude to DeepSeek and watched costs "crash to the ground."
-
AI News — June 23, 2026
Anthropic's Workload Identity Federation hit general availability — a workload can now swap its static `sk-ant-…` API key for short-lived tokens minted from an identity provider you already run (AWS, GCP, GitHub Actions, Okta…) and bound to per-workload service accounts that carry their own rate limits, usage attribution, and audit trail — while Claude Code v2.1.186 added a non-interactive `claude mcp login`, and Fable 5's no-extra-cost window closed with usage-credit metering starting today, the model still dark.
-
AI News — June 12, 2026
SpaceX priced the largest IPO in history — $75 billion raised at a $1.75 trillion valuation — carrying the merged xAI onto public markets first, while Anthropic published official terms for Monday's Agent SDK billing split.
-
AI News — June 11, 2026
OpenAI confirmed it has confidentially filed an S-1 with the SEC — one week after Anthropic — putting both leading labs on the IPO runway, while GitHub Copilot CLI gained a /security-review command.
-
AI News — June 9, 2026
Apple's WWDC keynote turned Apple Intelligence into a model marketplace — its new Extensions framework lets users set Claude, ChatGPT, Gemini, or Grok as the brain behind Siri and system AI this fall.
-
AI News — June 7, 2026
Cursor 3.7 pushes visual Design Mode into canvases and overhauls Teams pricing with a new Premium seat, while the Linux Foundation lines up a Tokenomics Foundation to standardize AI token-cost telemetry ahead of FinOps X.