On July 15, 2026, Thinking Machines Lab released Inkling, its first model: a 975-billion-parameter multimodal mixture-of-experts (about 41B active per query) trained from scratch on 45 trillion tokens of text, images, audio and video, under the permissive Apache 2.0 license. The company frames it around customization and ownership rather than topping benchmarks.
Mira Murati's Thinking Machines releases Inkling, an Apache-2.0 open model
If you need a model you can self-host, fine-tune and ship commercially without usage restrictions, an Apache-2.0 open-weight base like Inkling avoids per-token API lock-in; weigh infra cost against API convenience.
Source: TechCrunch
In the app, Founder Briefs are personalized to your country, industry and stage, and you can save the ones that matter.
More that helps you.
Anthropic ships Claude Opus 5 — near-top-tier performance at half the price
On July 24, 2026 Anthropic released Claude Opus 5, its fourth new model in under 60 days (after Sonnet 5, Fable 5 and Mythos 5 in June). Opus 5 delivers close to flagship Fable 5 q…
Google ships Gemini 3.6 Flash and 3.5 Flash-Lite at lower prices
On July 21, 2026, Google released Gemini 3.6 Flash plus 3.5 Flash-Lite and a gov-only Flash Cyber model, all with a 1M-token context window and day-one API/AI Studio availability.…
OpenAI launches GPT-5.6 family (Sol, Terra, Luna) with big efficiency gains
On July 9, 2026 OpenAI unveiled GPT-5.6 in three variants: Sol (the workhorse), Terra (mid-tier), and Luna (budget). Sam Altman says the models are markedly more token-efficient —…
Etched raises $300M at $10.3B valuation for transformer-only AI inference chips
On July 23, 2026 AI-chip startup Etched — founded by three Harvard dropouts in 2022 — closed a $300M Series C led by Sequoia at a $10.3B valuation, with a16z, Jane Street and SK Hy…