📜 The Primary Source Archive
30 briefings
OpenAI just pushed GPT-6 Astra into the market with a striking admission about its evasion behavior, formally triggering…
We are tracking a series of structural shifts in frontier AI economics and agent reliability this week. Anthropic's Fabl…
Today on The Primary Source: Anthropic's Fable 5.1 slashes cache-read costs by 75%, translating to massive per-task savi…
The gap between headline numbers and operational reality is widening this week. Sonnet 5's tokenizer change quietly push…
New benchmark data reveals the hidden cost multipliers in popular multi-agent frameworks, while the Mamdani administrati…
Anthropic is quietly clawing back Claude Code capacity under the guise of a permanent quota increase, while developers f…
Today on The Primary Source: Anthropic formally documents the limits of naive multi-agent scaling, an AWS engineering te…
Two architectural warnings land today: Microsoft published its agent governance spec just as Meta detailed the fallout o…
Today on The Primary Source: The MIT-licensed GLM-5.3-Flash fulfills the open-weights promise we've been tracking at one…
Today on The Primary Source: Anthropic patches a genuine cost-accounting gap for enterprise gateway operators, Moonshot'…
Two enterprise cost reports published today quantify an uncomfortable reality: cheaper per-unit inputs are driving overa…
Anthropic's latest patch notes just confirmed a silent, multi-month billing undercount for its U.S. inference tier, expo…
Today on The Primary Source: The frontier AI pricing structure is officially cracking. OpenAI is aggressively cutting AP…
The benchmarks that frontier AI labs rely on to market their models are quietly breaking under audit today. Elsewhere: U…
Standard AI pricing formulas are failing in production. Three independent measurements published today reveal that hidde…
Today on The Primary Source: We are seeing a widening gap between announced signals and operational reality. The Treasur…
Engineering teams are rolling out new optimizations for long-session agent costs today, directly targeting the prompt ca…
The API routing landscape shifted dramatically today as Stripe's $7 billion acquisition of OpenRouter merges model infer…
Cost structures for autonomous agents are shifting abruptly today. With DeepSeek enforcing time-of-day API multipliers a…
New benchmarks demonstrate that open-weights AI deployments can now match frontier models on offensive security tasks. I…
Inspectable agent harnesses and dynamic model routers dominate the engineering landscape today, as developers actively p…
Google and Anthropic formalized favorable API rates today, securing the unit economics necessary for autonomous agent wo…
Frontier AI labs are resetting the baseline economics for long-running agent workflows, as xAI and DeepSeek launch new p…
Today on The Primary Source, engineering teams are proving they can slash API spend by aggressively routing multi-turn t…
The mechanics of agent orchestration are hitting a new layer of maturity as abstract architecture proposals turn into pr…
DeepMind and MiniMax are squeezing multi-agent orchestration into tighter contexts today, marking the next step in the i…
The drive for sustainable AI unit economics accelerates today as DeepSeek pulls back its aggressive API discounts. Meanw…
Today on The Primary Source, Anthropic continues to reshape its developer pricing by doubling Claude Code rate limits, t…
The major AI labs are pushing forward with massive product updates this week, officially rolling out the models and agen…
A leadership shuffle at Google DeepMind and a major capability leak from Anthropic and OpenAI highlight the dual pressur…