
For developers building on @vercel.
San Francisco & Global
Joined November 2020
- Tweets1.4K
- Following5
- Followers49.9K
- Likes130
Pinned Tweet
Vercel Daemons is now in public beta.
Run background workers that pull jobs from queues, databases, or third-party systems.
vercel.com/changelog/vercel-β¦
AI Gateway and @eve now have free web search from Parallel.
@p0 search is now native to AI Gateway and the default π ππ_ππππππ for eve agents. Offer runs through January 2027.
vercel.com/changelog/paralleβ¦
You can now set a TTL on Vercel Blob writes.
Pass ππππ³ππ’π and the blob is deleted that many days after upload. Every write result and ππππ() response includes ππ‘ππππππ°π.
vercel.com/changelog/vercel-β¦
Daemons now supports OpenTelemetry export to Vercel Tracing.
Inspect worker spans alongside the rest of your project's traces.
vercel.com/changelog/vercel-β¦
Sandbox can now securely install your team's private npm packages.
Let agents build and test your app without exposing registry tokens.
vercel.com/changelog/vercel-β¦
Vercel Sandbox's Python SDK is now 1.0, with a stable API and all TypeScript SDK features.
β² ~/ uv add 'vercel-sandbox>=1.0'
vercel.com/changelog/vercel-β¦
1
28
1,922
New #1 on Deepsec Bench: GPT-6.2 Astra
It beats the last leader by 20 points. At 95% precision, it found nearly 50 more issues than any other model.
In second place, GPT-6.2 Sol beats the old leader by 6.6 points.
MAI-Image-2.7 and MAI-Image-2.7-Flash from @MicrosoftAI are live on AI Gateway with 2K output and transparent backgrounds. Try them:
ππ πππππ \
-π πππππππππ/πππ-πππππ-πΈ.π½ \
"π°π πππππππππ"
Decision-1.1 from @MicrosoftAI is live on AI Gateway. It is more accurate, better calibrated, and faster than Decision-1.
πππππππππ/ππππππππ-π·.π·
vercel.com/changelog/decisioβ¦
GPT-6.2 Astra and Sol are live on AI Gateway. Better output with far fewer tokens, and a lower price on Astra.
ππππππ/πππ-πΌ.πΈ-πππππ
ππππππ/πππ-πΌ.πΈ-πππ
vercel.com/changelog/gpt-6-2β¦
You can now set budgets for individual projects and products on Vercel.
vercel.com/changelog/set-budβ¦
Hy4 Flash from @TencentHunyuan is live on AI Gateway today.
β’ Hy4's capabilities at 1/10th the price
β’ Only 12B activated params
β’ Additional 50% off through Nov 29
πππππππ/ππ’πΊ-πππππ
vercel.com/changelog/hy4-flaβ¦
5
32
2,994
Qwen4-Max from @Alibaba_Qwen is now available on AI Gateway.
β’ Built for 3D modeling and game dev
β’ 25% lower pricing than Qwen3.8-Max
β’ Up to 384K output tokens
πππππππ/ππ πππΊ-πππ‘
vercel.com/changelog/qwen4-mβ¦
fx v0.0.17
Subagents & Token Pass, first-class.
β’ Steer subagents as they work
β’ Token Pass models built in
β’ 48.0% smaller tool definitions
fx.sh/changelog#v0.0.17
Vercel Token Pass now offers $150 in tokens for $30 per month.
β’ Use across the 5 models supported today
β’ Available in fx with no extra setup
vercel.com/changelog/vercel-β¦
Vercel Sandbox now starts up to 4x faster.
Images are now read from regional caches when a sandbox starts, making startup fast in every Vercel region β
vercel.com/changelog/vercel-β¦
Vercel Security Review no longer requires a Vercel project to scan your GitHub repos.
Every push to a connected repo gets code changes reviewed for vulnerabilities β
vercel.com/changelog/vercel-β¦
Every build machine on Vercel now gets a 5 GB build cache.
Fewer cache misses on large projects means faster installs and builds β
vercel.com/changelog/build-cβ¦
Parallel Decisions from @p0 is on AI Gateway, free until Nov 23.
A decision model that searches the web, returns a probability, and lists the citations alongside it.
ππππππππ/ππππππππ-πππππππππ-ππππ
vercel.com/changelog/paralleβ¦

