Web Pulse — AI & Tech News for Agents

agent @ wpnews : ~/news   $ tail -f --follow=name /var/news/latest
[LATEST] src=byteiota.com pub= verified=true topic=ai-chips

OpenAI’s Jalapeño Chip Benchmarks: What Developers Need to Know

OpenAI unveiled benchmark results for its Jalapeño inference chip at Hot Chips 2026, claiming 1.7x to 3.6x lower end-to-end latency and 1.5x to 1.9x better work per watt than Nvidia's GB300 at half the power draw (700W vs 1,400W). The chip, co-developed with Broadcom on TSMC 3nm, is not for sale but will power all OpenAI API calls eventually, with Microsoft taking roughly 40% of initial production…

#openai #broadcom #tsmc #nvidia #microsoft #semianalysis
~/api · curl 200 · 24ms · application/json
// GET /v1/news → latest 10 articles
$ curl https://api.wpnews.pro/v1/news?limit=10 
[
  { "id": "art-001", "src": "byteiota.com",
    "title": "OpenAI’s Jalapeño Chip Benchmarks: What Develope…",
    "published": "2026-08-30T00:10:08+00:00",
    "sentiment": "neutral" },
  // … 9 more
]
JSON MCP OpenAI tools RSS webhooks
→ /for-agents → /api-docs → /integrations
── latest · today's filings 10 results · sorted by recency
all artificial intelligence ai products ai infrastructure ai tools ai agents developer tools ai policy large language models all topics →
more stories 6 below
23:53
2026-08-29
simonwillison.net
large-language-models · 1m read · neu

Introducing Hy4 Preview

Tencent released Hy4 Preview, a large language model with a significant size increase from its predecessor Hy3, which had 295B total parameters, 21B active parameters, 256,000 context length, and 598GB size. The new mode…

tencent hy4-preview hy3 hugging-face openrouter
$ load --more ↓
── trending today this week
  1. 1 Google begins rolling out AI-focused homepage, ditches search button to make room for tools like image generation 129 views
  2. 2 The mystery model testing AI’s shrinking moats 128 views
  3. 3 I built an agent-skill that turns your texts into diagrams and BI-visuals 115 views
  4. 4 Kimi K3: The Open-Weight Frontier Model Devs Should Know 115 views
  5. 5 The Best Open LLM You Can Actually Run Right Now, by VRAM Tier (August 2026) 115 views
── in the news entities · 30d
OpenAI 946 Anthropic 811 Google 551 Claude 419 Claude Code 390 Nvidia 371 Microsoft 325 Hugging Face 284 all index →
── hot today ranked by views
  1. 1 LocalAI – Run any model – LLMs, vision, voice, image, video – on any hardware 67
  2. 2 Aider: AI pair programming ใน terminal — เทียบหมัดต่อหมัดกับคู่แข่ง 5 ตัว 60
  3. 3 Quanta.Ai.Code.Editor 55
  4. 4 What is MCP Protocol? The USB Port for AI Agents 54
  5. 5 Building CareLoop: an autonomous clinical-triage agent where rules decide and AI explains 53
── today's tech-news mood auto-classified
10.0% positive, 78.0% neutral, 12.0% negative.
positive 10.0% neutral 78.0% negative 12.0%
── topic momentum (7d) gainers
Mlops +22%
Ai Chips +12%
── the archive browse · compare · explore
Daily Digest
Browse articles by date
Today's Digest →
Weekly Roundup
Top AI news of the week
This Week →
Full Archive
All articles by month
Browse Archive →
Coverage Check
Who dominates AI news?
AI Index →
Compare
ChatGPT vs Claude vs Gemini
Compare →
LLM Models
GPT-4o vs Claude 3.5 vs Gemini 1.5
LLM Ranking →
For developers & agents

Free JSON API · 75 OpenAI tools · 65 MCP tools for Claude

Plug into Claude, ChatGPT, or any agent framework. JWT-scoped keys for prod, no rate limit for dev. Structured entity / topic metadata on every article.

$ curl -H "Authorization: Bearer $WPN_KEY" https://api.wpnews.pro/v1/news
uptime 99.97% latency p50 24ms updated every 30 min free tier 10k req/day
Get free key → More patterns → MCP server RSS · Atom
sponsored brought to you by zahid.host 4,200+ projects deployed · EU only
welcome back — you have 3 new commits ready to deploy.

Ship your project in 30 seconds with zahid.host

EU-based hosting built for the people who actually read this site: developers, indie hackers, and AI agents shipping side projects. Git-push deploys, automatic HTTPS, fair pricing.

--git-deploy  push to deploy
--eu-region  Kyïv / Warsaw / Frankfurt
--uptime    99.99% SLA
--https-auto Let's Encrypt
--no-cold   always-warm runtimes
--free-tier custom domain incl.
$git push zahid main
→ Authenticating… ✓
→ Building… ✓ (12.4s)
→ Live at https://your-app.zahid.host ✓
Get free account → See pricing Deploy docs
from €0/mo · scale on demand
no card required · cancel with one curl
/// newsletter

Stay sharp.

The best AI & dev news, curated daily and piped straight to your inbox. No spam. Unsubscribe with one curl.

── emerging this week topics with the biggest jumps
artificial intelligence ai products ai infrastructure ai tools ai agents developer tools ai policy large language models ai research ai safety ai ethics generative ai ai startups machine learning ai chips natural language processing all topics →
LIVE [news] indexed:106136 sources:3030 synced: 2026-08-30 · 00:10:08