Thinking Machines launches Inkling 2 Small AI model
The company says the model performs better at agentic tasks, but that gain comes at a higher token cost.
By Standard Topic. AI-generated from the reporting listed under Coverage. Published . Updated .
This story is no longer being updated. It reflects reporting up to .
What happened
Thinking Machines is releasing Inkling 2 Small, a mixture-of-experts model with 594 billion parameters. The company says it is the strongest western open-weight model, though its token usage rises sharply. An analysis by Vals AI found that the model swears during its internal reasoning.
Why it matters
The company says the model performs better at agentic tasks, but that gain comes at a higher token cost. The profanity finding raises questions about how reliably its reasoning signals failure.
Timeline
- Nov 6, 8:25 AM ET: Thinking Machines announced the release of Inkling 2 Small, a 594-billion-parameter mixture-of-experts model.
- Nov 6, 4:27 PM ET: Reports said the company was pursuing a $5 billion fundraising push alongside the release.
- Nov 7, 11:31 AM ET: An analysis by Vals AI found the model swears during its internal reasoning.
Coverage
- Thinking Machines' Inkling 2 Small drops 316 F-bombs across benchmark runs (the-decoder.com)
- Thinking Machines debuts Inkling 2 Small amid reported $5B fundraising push (siliconangle.com)
More from Technology & AI
- China puts humanoid robots to work serving food and sorting parcels
- Newsom proposes sweeping AI safety rules for Washington
- Google releases Gemini 4 Neon AI model with million-token context
- Microsoft and Intel widen their cloud chip partnership
- Google DeepMind rolls out Genie 4 to select subscribers
- AI helped AMD shorten chip design work, OpenAI says