Claude Fable 5 vs Opus 4.8: What Creators Get
Anthropic released Claude Fable 5, its most capable model, on June 9, 2026, free on paid plans through June 22. How it compares to Opus 4.8.
Xiaomi MiMo Hits 1000 Tokens Per Second on 1T Open Model
Xiaomi MiMo-V2.5-Pro-UltraSpeed claims 1000 tokens per second decode on a 1T MoE. Open-weights FP4 checkpoint plus a 2-week free API trial.
ICML 2026: AI Models Could Run on 97% Less Memory
New ICML 2026 research shows transformer models can share attention projections, achieving up to 96.9% KV cache reduction with minimal accuracy loss.
Nemotron 3 Ultra: NVIDIA's 550B Open-Weights MoE
NVIDIA released Nemotron 3 Ultra on June 1 2026: a 550B mixture-of-experts model with 55B active parameters, open weights on Hugging Face, with 5x faster inference and 30% lower cost than Nemotron 2.
Tesla V100 in a Gaming PC: 32GB Local LLM for £200
Developer Oscar Molnar installed a secondhand Tesla V100 SXM2 into his gaming PC alongside an RTX 4080, building a 32GB dual-GPU setup for under £200 total.
George Hotz: AI Agents Are a Slot Machine for Creative Work
George Hotz argues AI agents frontload impressive progress but stall on polish, creating a golden era of AI-generated slop that creators need to plan for.
Qwen3.7-Max vs Claude, Gemini, GPT-5.5: Compared
Qwen3.7-Max sets a new bar for non-hallucination rate among frontier agent models. Here is how it stacks up against Claude Opus 4.7, Gemini 3.1 Pro, and GPT-5.5 across the four reliability dimensions that decide which model goes into production.
Cohere Command A+ Opens 218B Apache 2.0 Frontier Model
Cohere released Command A+ under Apache 2.0 on May 21, 2026. The 218B sparse MoE runs on two H100s, with native citations and 48 languages.
NVIDIA Nemotron Diffusion: 3x Faster LLM Decoding
NVIDIA released the Nemotron-Labs-Diffusion family on Hugging Face, an open-weights LLM that switches between autoregressive, diffusion, and self-speculation decoding for 2.7x to 3.3x throughput gains.
Gemini 3.5 Flash: $1.50 Coding Model Beats 3.1 Pro
Google launched Gemini 3.5 Flash at I/O 2026, a Flash-tier model that beats 3.1 Pro on coding and agent benchmarks at 40 percent lower cost.
Qwen 3.7 Max and Plus Preview Land in Arena Top 15
Alibaba pushed Qwen 3.7 Max Preview and Qwen 3.7 Plus Preview to Arena and Qwen Chat for testing. Max sits 13th overall on Arena Text and Plus is 16th on Vision.
DeepSeek-V4-Flash Makes LLM Steering Practical
DeepSeek-V4-Flash is the first local model competitive with frontier AI, making LLM activation steering practical for the first time. A guide for creators.
RelaxAI: UK-Hosted LLM API 80% Cheaper Than OpenAI
Civo Limited’s RelaxAI offers UK-sovereign LLM inference at £0.10 per million input tokens, with ISO 27001 certification and 100% UK data residency.
OpenAI vs. Apple: The ChatGPT-Siri Deal Is Falling Apart
OpenAI enlisted outside legal counsel to explore action against Apple after the ChatGPT-Siri integration failed to generate expected subscription revenue.
OpenAI Codex Now in ChatGPT iOS: Monitor Projects Anywhere
OpenAI Codex is now in the ChatGPT iOS app, letting creators monitor projects, review diffs, and push code changes without being at a desk.
Recursive Superintelligence: $650M for Self-Improving AI
Richard Socher's stealth-mode AI lab Recursive exits with $650M at a $4.65B valuation, backed by GV, Greycroft, Nvidia, and AMD.
Anthropic Now Has More Business Customers Than OpenAI
For the first time, more businesses pay for Anthropic than OpenAI, according to May 2026 spending data from Ramp showing 34.4% vs 32.3% business adoption.
Mistral Python Package Backdoor: Check Your SDK Now
mistralai==2.4.6 was backdoored in the Mini Shai-Hulud supply chain attack. PyPI quarantined the project. Here is what to check and do if you build with Mistral's Python SDK.
Thinking Machines TML-Interaction: Full-Duplex Voice AI
Mira Murati's Thinking Machines shipped TML-Interaction-Small, a 276B full-duplex voice model that listens and speaks simultaneously, beating GPT-Realtime on latency and interaction quality.
Claude Agents Gain Self-Learning Memory and Parallel Teams
Anthropic adds four capabilities to Claude Managed Agents: dreaming (self-improving memory), outcomes (quality grading), multiagent orchestration, and webhooks. Harvey saw a 6x completion improvement; Wisedocs cut review cycles by 50%.
SubQ Launches With 12M Context and 300x Lower Cost
Subquadratic SubQ launched May 5 with a 12M token context window and an OpenAI-compatible API. Headline claim: 300x lower cost than Claude Opus on RULER 128K.
GPT-5.5 Instant: ChatGPT's New Default Cuts Hallucinations
OpenAI replaced GPT-5.3 Instant with GPT-5.5 Instant as ChatGPT's default model, citing 52% fewer hallucinated claims and a new memory source viewer.
iOS 27 Brings AI Model Choice to Apple Intelligence
iOS 27 Extensions lets you swap in Claude, Gemini, ChatGPT, or Grok for Apple Intelligence tasks (Writing Tools, Image Playground, Siri). Comparison + setup workflow + pricing math for working creators using iPhone, iPad, and Mac in 2026.
Anthropic's $50B Round Targets a Compute Ceiling
Anthropic is closing a $50 billion round at an $850 to $900 billion valuation, the largest private AI raise in history and the company's final financing before a late-2026 IPO.