Gawk · 2026-05-19
Gawk — 2026-05-19 · 5 tool incidents
Five verifiable things that moved in the AI ecosystem in the last 24h. Every number traces to a public source.
What moved
- · claude-opus-4-6-thinking holds #1 on LMArena for the 28th consecutive snapshot.
- · ai downloads grew for the 5th consecutive snapshot.
- · openai downloads grew for the 5th consecutive snapshot.
Benchmark movers
#No rank changes in the LMArena top 3
Why this matters · Public benchmarks are gameable, but rank shuffles still hint at where the frontier is genuinely moving.
#1 claude-opus-4-6-thinking
anthropic · 1500 Elo
#2 claude-opus-4-6
anthropic · 1498 Elo
#3 claude-opus-4-7-thinking
anthropic · 1485 Elo
Source: lmarena.ai
Tool Health
#5 incidents in the last 24h (vs 1 yesterday)
Why this matters · Provider outages and degradations cause retry storms upstream. Tracking the 7-day shape catches flapping providers before they page you.
Elevated errors on Claude Opus 4.7
started 08:41 UTC · resolved 10:08 UTC · minor impact
Elevated errors on Claude Haiku 4.5
started 07:00 UTC · resolved 08:30 UTC · minor impact
Elevated errors on Claude.ai and Claude Code login
started 04:46 UTC · resolved 04:51 UTC · minor impact
Elevated errors for Claude Opus 4.7
started 22:10 UTC · resolved 22:19 UTC · major impact
Windsurf Authentication Slowness
started 13:22 UTC · resolved 14:40 UTC
Source: status.anthropic.com, status.windsurf.com
Top HN stories
#Top 5 on HN in the last 24h
Why this matters · What developers debate on Hacker News often previews which models, frameworks, and patterns they'll actually adopt next.
Elon Musk has lost his lawsuit against Sam Altman and OpenAI
112 points · 36 comments
The last six months in LLMs in five minutes
79 points · 22 comments
What political censorship looks like inside an LLM's weights (Qwen 3.5)
71 points · 32 comments
Mexican government breached by solo user with Claude, 150 GB exfiltrated
44 points · 35 comments
Cursor Introduces Composer 2.5
29 points · 7 comments
Source: news.ycombinator.com
SDK adoption
#6 notable shifts across the six registries
Why this matters · Package install volume is where developers place real bets — distinct from where labs are shipping marketing.
PyPI: openai
+4.1M 24h downloads day-over-day
pypistats.org is a third-party aggregator — counts include mirror hits, CI builds, and pip install retries.
npm: openai
+44.6k 24h downloads day-over-day
crates.io: candle-core
+13.7k 90d downloads day-over-day
Docker Hub: ollama/ollama
+333.1k all-time pulls day-over-day
Homebrew: ollama
−718 30d downloads day-over-day
VS Code Marketplace: GitHub.copilot
+9.0k cumulative installs day-over-day
Source: pypistats.org, npmjs.com, crates.io, hub.docker.com, formulae.brew.sh, marketplace.visualstudio.com
Agent frameworks
#3 agent frameworks moved >10% in the last week
Why this matters · Agent frameworks rise and fall on weekly download velocity and open-issue trajectory, not GitHub stars.
CrewAI · +82%
4.0M weekly downloads · 52k stars
smolagents · +18%
167k weekly downloads · 27k stars
OpenAI Agents · +17%
9.2M weekly downloads · 26k stars
Source: pypistats.org, pypistats.org, pypistats.org
Notable AI Lab activity
#10 labs moved in the last 24h
Why this matters · GitHub event volume on a lab's own repos is the cleanest publicly-verifiable proxy for engineering activity we have.
Anthropic
+15 events (now 200) · San Francisco, US
LangChain
+5 events (now 195) · San Francisco, US
OpenAI
+45 events (now 168) · San Francisco, US
Hugging Face
+19 events (now 101) · New York, US
Microsoft Research
−5 events (now 87) · Redmond, US
DeepSeek
−17 events (now 65) · Hangzhou, CN
New mover: Weights & Biases
59 events · San Francisco, US
New mover: Replicate
30 events · San Francisco, US
Dropped off: xAI
was 27 events · Palo Alto, US
Dropped off: INRIA
was 23 events · Paris, FR
Source: gawk.dev
Model Usage
#tencent/hy3-preview holds #1 · biggest mover: z-ai/glm-4.7 +16 to #30 · biggest drop: z-ai/glm-5 -10
Why this matters · OpenRouter rankings reflect API-first developer spend; direct customers like consumer ChatGPT are invisible by construction.
z-ai/glm-4.7 climbed +16 to #30
was #46 on 2026-05-12
OpenRouter request volume reflects developer API spending, not end-user adoption. Biased toward API-first workflows that route through OpenRouter — direct OpenAI / Anthropic / Google customers who never use OpenRouter are invisible.
z-ai/glm-5 slipped -10 to #39
was #29 on 2026-05-12
#1 tencent/hy3-preview
#2 deepseek/deepseek-v4-flash
#3 anthropic/claude-sonnet-4.6
Source: openrouter.ai