Cursor Composer 2.5 ships as a Cursor-only ~1T-param model trained on xAI infrastructure, ~10x cheaper on-platform, amid acquisition talks post-IPO.
SpaceX AI closes $60B Cursor acquisition post-IPO; Anthropic simultaneously pays $1.25B/month for Colossus 2 compute through May 2029 — a structural conflict with no AI-industry precedent.
xAI releases Grok Build for SuperGrok subscribers — terminal coding agent with Plan Mode, parallel subagents, CI headless mode, and native CLAUDE.md support.
xAI's Grok Build adds a 2M-token context and Plan Mode to the terminal-agent race — competing with Claude Code and Codex CLI, locked behind SuperGrok Heavy at $300/month.
Grok downloads drop from 20M+ in January to 8.3M in April 2026 — a 60%+ decline — while paid US adoption is flat year-over-year per Recon Analytics.
Devendra Chaplot — Mistral AI founding member and marquee xAI hire — has departed after just one month, raising questions about talent retention at xAI.
Anthropic secures all of SpaceX's Colossus 1 (220K GPUs, 300 MW) and doubles Claude Code rate limits—sub-agent fan-out is now viable at $200/month.
xAI's Grok 4.3 launches with 1M-token context, native video input, always-on reasoning, and the most aggressive long-context pricing in the frontier tier.
Anthropic leases all of SpaceX Colossus 1 (300 MW, 220K GPUs), doubles Claude Code limits, and ships Managed Agents, Dreaming, and Outcomes in one day.
xAI's new voice cloning API offers 80+ voices, 28 languages, and 2-min training at $3/hr. Custom voices unlock full TTS with identity verification.
Independent benchmarks reveal Grok 4.3 beta trails multiple open competitors, while Mistral Medium 3.5 underperforms DeepSeek V4 at higher cost on Artificial Analysis.
Eight AI companies signed 'any lawful purpose' DoD agreements on May 1 — Anthropic was excluded after a guardrail dispute, while the NSA quietly continued using Claude Mythos anyway.
Elon Musk's $134B lawsuit against OpenAI and Sam Altman begins federal trial April 27 in Oakland, placing OpenAI's nonprofit-to-for-profit conversion under judicial scrutiny.
xAI's Grok TTS API achieves 5% word error rate vs ElevenLabs' 12%, now available via API at below-ElevenLabs pricing — a direct challenge to the leading voice AI provider.
SpaceX S-1 discloses multiple CSAM investigations into xAI's Grok as a material risk to global market access — a first for an IPO filing.
SpaceX AI pairs its Colossus compute cluster with Cursor's coding-agent demand in a deal valuing Cursor at $60B — xAI gets demand, Cursor gets frontier-model access.
Grok-3 brings vision, multi-step reasoning, and real-time X platform data integration to xAI's model lineup.
Curated AI insights — sent when there's something worth your inbox.