Gawk · 2026-06-17
Gawk — 2026-06-17 · 7 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 30th consecutive snapshot.
- · ai downloads grew for the 5th consecutive snapshot.
- · transformers downloads grew for the 5th consecutive snapshot.
Benchmark movers
#1 change in the LMArena top 3
Why this matters · Public benchmarks are gameable, but rank shuffles still hint at where the frontier is genuinely moving.
claude-opus-4-6-thinking
+10 Elo (now 1511) · anthropic
Source: lmarena.ai
Tool Health
#7 incidents in the last 24h (vs 4 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 for Claude Opus 4.8
started 08:24 UTC · resolved 10:03 UTC · minor impact
Elevated errors for Claude Opus 4.8
started 05:08 UTC · resolved 06:37 UTC · minor impact
Elevated errors on Claude Sonnet 4.6 and Opus 4.8
started 00:47 UTC · resolved 02:50 UTC · minor impact
Elevated errors for Claude Opus 4.8
started 20:50 UTC · resolved 22:02 UTC · major impact
Opus 4.8 errors
started 19:59 UTC · resolved 20:16 UTC
Elevated errors across many models
started 17:29 UTC · resolved 19:32 UTC · minor impact
Incident with Copilot Availability
started 03:50 UTC · resolved 04:44 UTC · critical impact
claude-code: Operational → Degraded
Source: status.anthropic.com, githubstatus.com, status.claude-code.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.
Leaked OpenAI financials show $38.5B loss and compute burn
103 points · 79 comments
SpaceX to buy Cursor AI coding agent operator Anysphere for $60B
94 points · 45 comments
Claude: Elevated errors across many models
15 points · 0 comments
Show HN: Building Your Own Coding Agent on Top of Zot – Meet Coil
9 points · 0 comments
SpaceX to buy Cursor for $60B (WSJ)
7 points · 1 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: langchain
−228.1k 24h downloads day-over-day
pypistats.org is a third-party aggregator — counts include mirror hits, CI builds, and pip install retries.
npm: openai
+2.0M 24h downloads day-over-day
crates.io: ort
+34.6k 90d downloads day-over-day
Docker Hub: ollama/ollama
+281.7k all-time pulls day-over-day
Homebrew: ollama
−365 30d downloads day-over-day
VS Code Marketplace: GitHub.copilot
+12.9k cumulative installs day-over-day
Source: pypistats.org, npmjs.com, crates.io, hub.docker.com, formulae.brew.sh, marketplace.visualstudio.com
Agent frameworks
#4 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.
Pydantic AI · +68%
6.2M weekly downloads · 18k 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.
OpenAI Agents · -12%
7.1M weekly downloads · 27k stars
AutoGen · -11%
288k weekly downloads · 59k stars
pypistats.org · View on Gawk →
Tracks the live `autogen-agentchat` package — the deprecated bare `autogen` PyPI name is a different abandoned project and is not counted here.
CrewAI · -11%
2.4M weekly downloads · 54k stars
Source: pypistats.org, pypistats.org, pypistats.org, pypistats.org
Notable AI Lab activity
#9 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
+16 events (now 194) · San Francisco, US
LangChain
−9 events (now 188) · San Francisco, US
New mover: Mistral AI
185 events · Paris, FR
Microsoft Research
+13 events (now 166) · Redmond, US
OpenAI
−13 events (now 141) · San Francisco, US
Replicate
−8 events (now 81) · San Francisco, US
Weights & Biases
−9 events (now 70) · San Francisco, US
DeepSeek
−9 events (now 54) · Hangzhou, CN
Dropped off: Intel AI
was 26 events · Santa Clara, US
Source: gawk.dev
Model Usage
#z-ai/glm-5.2 holds #1 · biggest mover: qwen/qwen3.7-max +24 to #15 · biggest drop: tencent/hy3-preview -45
Why this matters · OpenRouter rankings reflect API-first developer spend; direct customers like consumer ChatGPT are invisible by construction.
qwen/qwen3.7-max climbed +24 to #15
was #39 on 2026-06-09
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.
tencent/hy3-preview slipped -45 to #47
was #2 on 2026-06-09
#1 z-ai/glm-5.2
#2 openrouter/fusion
#3 moonshotai/kimi-k2.7-code
Source: openrouter.ai