Blog

Insights, tutorials, and updates from the ZharfAI team

News

Major AI news and what it means for business and markets, written from primary sources: what was announced, what we checked and what is still unconfirmed.

OpenAI Calls Astra Cyber-Critical and Limits Its Release
NewsSeptember 2, 2026•9 min read

OpenAI Calls Astra Cyber-Critical and Limits Its Release

OpenAI says its unreleased Astra model is the first it has designated at the Critical cybersecurity threshold.

OpenAI Puts ChatGPT Ads at a $1 Billion Annualized Run Rate
NewsSeptember 1, 2026•8 min read

OpenAI Puts ChatGPT Ads at a $1 Billion Annualized Run Rate

OpenAI says ChatGPT Ads reached a $1 billion annualized revenue run rate in under 200 days; the figure is company-reported, not audited revenue.

llama.cpp Cuts Redundant Scans, Lifting One Long-Context Test 51%
NewsAugust 31, 2026•7 min read

llama.cpp Cuts Redundant Scans, Lifting One Long-Context Test 51%

llama.cpp release b10707 stops an n-gram lookup from scanning all 256 possible sequence slots after it has found every sequence actually attached to a used KV-cache cell.

llama.cpp Makes Adreno Prefill Up to 25% Faster Without New Silicon
NewsAugust 30, 2026•7 min read

llama.cpp Makes Adreno Prefill Up to 25% Faster Without New Silicon

llama.cpp release b10687 changes how two Qualcomm Adreno GPU generations route matrix multiplication.

OpenAI Adds Compute Units to Its SDK, but the Meter Has No Number Yet
NewsAugust 29, 2026•7 min read

OpenAI Adds Compute Units to Its SDK, but the Meter Has No Number Yet

OpenAI’s Python SDK 3.6.0 adds an optional compute_units field to usage objects for Responses and Chat Completions.

AI Capacity Splits Into Tiers; Contracts Carry the Capital
NewsAugust 28, 2026•8 min read

AI Capacity Splits Into Tiers; Contracts Carry the Capital

vLLM 0.28.0 adds disk-backed tiers for the inference KV cache, while IREN says its 2026 capacity is largely sold out under $4 billion of contracted run-rate revenue.

NVIDIA Turns AI Demand Into Cash; Korea Raises Capital’s Price
NewsAugust 27, 2026•8 min read

NVIDIA Turns AI Demand Into Cash; Korea Raises Capital’s Price

NVIDIA reported $96.2 billion of quarterly revenue and $74.4 billion of first-half operating cash flow; hours later, the Bank of Korea raised its policy rate.

AI Deployment Gets Wider; Earnings Quality Becomes the Test
NewsAugust 26, 2026•8 min read

AI Deployment Gets Wider; Earnings Quality Becomes the Test

New llama.cpp and MLX releases reached harder deployment details, while Zoom's results showed why earnings quality, not headline profit, is the test.

WebGPU Opens the Browser Edge; Alibaba Prices AI's Capital Burden
NewsAugust 25, 2026•8 min read

WebGPU Opens the Browser Edge; Alibaba Prices AI's Capital Burden

Microsoft's ONNX Runtime WebGPU Plugin EP 0.3.0 widened the practical route for generative models to run through WebGPU.

AI Inference Finds a Shortcut; Energy Sets the Price Floor
NewsAugust 24, 2026•7 min read

AI Inference Finds a Shortcut; Energy Sets the Price Floor

llama.cpp added multi-token prediction for GLM-4.5-Air, lifting one contributor benchmark from 74.55 to 88.73 tokens per second, as Singapore's power prices jumped.

AI Agents Learn to Resume; Jobs and Credit Stay Uneven
NewsAugust 23, 2026•7 min read

AI Agents Learn to Resume; Jobs and Credit Stay Uneven

Microsoft's stable Agent Framework releases for Python and .NET move agent continuity out of process memory and into explicit infrastructure.

AI Demand Clears the Proof Bar; Access and Cash Stay Gated
NewsAugust 22, 2026•7 min read

AI Demand Clears the Proof Bar; Access and Cash Stay Gated

Anthropic extends Claude Mythos 5 cyberdefense under vetted access and human approval, and Alibaba reports AI Cloud and Compute revenue up 45%.