Platform Profile
Thinking Machines
Mira Murati’s lab building open-weight multimodal base models and the Tinker customization platform—betting enterprises want models they can fine-tune, not only rent.
Role
Open-weight customization
Stack
MoE transformers, Tinker fine-tuning, multimodal RL
Status
Active
Thinking Machines Lab, founded by former OpenAI CTO Mira Murati, trains open-weight foundation models (Inkling family) and operates Tinker, a platform for fine-tuning and deploying customized checkpoints across inference partners.
Related Coverage
- Thinking Machines Releases Inkling-Small Open Weights
On July 30, 2026, Thinking Machines Lab released Inkling-Small—a 276B-total / 12B-active open-weight MoE that beats full Inkling (975B/41B) on SWE-Bench Verified (80.2% vs 77.6%) and ARC-AGI-2 (40.1% vs 36.5%) at roughly one-quarter the size and $0.30/$1.20 per million tokens.
- Mira Murati’s Thinking Machines Releases Inkling, Its First Open-Weight Model
On July 15, 2026, Thinking Machines Lab—founded by former OpenAI CTO Mira Murati—released Inkling, a 975B-parameter open-weight MoE multimodal model (41B active) with a 1M-token context window, full weights on Hugging Face, and fine-tuning via the company’s Tinker platform.