Creeta — AI developer tools & ecosystem news
v0.22.0: the chips sat idle while the front end was choking
vLLM v0.22.0 ships --api-server-count, a DP Supervisor, and three LB topology modes. Annotated explainer for operators.
Mistral's chip ambition: conditional. Its EU cluster: 44 MW.
What Mensch said about Mistral chips, and what's confirmed: EU cluster specs, ASML deal, and the sovereign compute bet.
Meta Business AI went global — gated rollout, paid plans TBD
Meta Business AI: activation flow, four controls, WhatsApp messaging fees, and the fine print on a gated market rollout.
178 desk rejections on a parameter authors never saw
NeurIPS 2026 desk-rejected 18% of position papers via Pangram 3.3.2 — methodology, calibration, June 15 deadline.
SynthID runs in ChatGPT. A blank result proves nothing.
Two provenance layers now cover Search, Gemini, Chrome, and Pixel. What each catches, where both fail, and what the new Cloud detection interface gives developers.
CMG sold email lists. They called it AI voice targeting.
FTC's $930K proposed consent order against Cox Media Group exposes how 'Active Listening' AI ad targeting was repackaged email lists — and why buried ToS can't substitute for voice-data consent.
MiniMax M3 benchmarks at $0.30/M: verified vs. vendor-only
MiniMax M3 at $0.30/M: what the 1M-sequence benchmarks mean, credential selection, and a quickstart.
181 Firefox exploits, no mandatory submission: Trump's AI EO
Trump's June 2026 EO sets a voluntary 30-day AI cyber review. What labs submit, who grades it, and what Mythos proved.
MAI Is Already in Your IDE — the Flagship Is Still Gated
Seven MAI models at Build 2026: what's live in Copilot, what's gated, and what Microsoft's technical report claims — all vendor-reported until verified.
NVIDIA's 550B finally lands: free to use, expensive to host
Nemotron 3 Ultra, 550B MoE (June 4 2026): hardware minimums, hosted API quickstart, NIM steps, benchmark check.
MAI-Thinking-1 beats Anthropic's top model — per Microsoft
Seven MAI models at Build 2026. MAI-Thinking-1 is a 35B-active sparse MoE — specs, claimed scores, and what's still unverified externally.
Qwen3 in the browser, zero keys — WebLLM 0.2.83 hands-on
WebLLM 0.2.83: run Qwen3 in Chrome via WebGPU, no server. Setup steps, streaming code, VRAM requirements, and gotchas.
AI developer tools and ecosystem news for developers and technical founders