Gawk · 2026-05-20
Gawk — 2026-05-20 · 3 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 29th consecutive snapshot.
- · ai downloads grew for the 6th consecutive snapshot.
- · openai downloads grew for the 6th 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
#3 incidents in the last 24h (vs 5 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 Haiku 4.5
started 08:14 UTC · resolved 08:49 UTC · minor impact
Elevated errors on Claude Opus 4.7
started 14:03 UTC · resolved 15:40 UTC · major impact
API users may see increased error rates for GPT-5.4 and GPT-5.5
started 23:32 UTC · resolved 00:37 UTC · minor impact
Source: status.anthropic.com, status.openai.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.
OpenAI Adopts Google's SynthID Watermark for AI Images with Verification Tool
55 points · 24 comments
Anthropic Is Preparing for IPO and We Should Be Worried
30 points · 30 comments
Show HN: Chrome ext to let zot, your terminal coding agent, operate the browser
11 points · 1 comments
We cut Claude's token usage 79% by redesigning our CLI for agents
11 points · 3 comments
Andrej Karpathy Joins Anthropic
9 points · 0 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
+734.9k 24h downloads day-over-day
pypistats.org is a third-party aggregator — counts include mirror hits, CI builds, and pip install retries.
npm: openai
+1.8M 24h downloads day-over-day
crates.io: candle-core
+5.8k 90d downloads day-over-day
Docker Hub: ollama/ollama
+341.7k all-time pulls day-over-day
Homebrew: ollama
−2.6k 30d downloads day-over-day
VS Code Marketplace: GitHub.copilot
+28.5k cumulative installs day-over-day
Source: pypistats.org, npmjs.com, crates.io, hub.docker.com, formulae.brew.sh, marketplace.visualstudio.com
Agent frameworks
#2 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 · +81%
4.0M weekly downloads · 52k stars
Pydantic AI · -33%
8.0M weekly downloads · 17k stars
pypistats.org · View on Gawk →
Tracks the `pydantic-ai` meta-package; the lean `pydantic-ai-slim` variant is roughly 4× larger but users typically compare the meta number.
Source: 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.
LangChain
−24 events (now 171) · San Francisco, US
OpenAI
−33 events (now 135) · San Francisco, US
Microsoft Research
+31 events (now 118) · Redmond, US
Hugging Face
−8 events (now 93) · New York, US
New mover: Google Research
92 events · Mountain View, US
Ollama
−8 events (now 89) · San Francisco, US
DeepSeek
+17 events (now 82) · Hangzhou, CN
New mover: AMD
66 events · Santa Clara, US
Dropped off: Replicate
was 30 events · San Francisco, US
Dropped off: Meta FAIR
was 20 events · Menlo Park, US
Source: gawk.dev
Model Usage
#tencent/hy3-preview holds #1 · biggest mover: z-ai/glm-4.7 +12 to #30 · biggest drop: z-ai/glm-5 -8
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 +12 to #30
was #42 on 2026-05-13
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 -8 to #39
was #31 on 2026-05-13
#1 tencent/hy3-preview
#2 deepseek/deepseek-v4-flash
#3 anthropic/claude-sonnet-4.6
Source: openrouter.ai