Creators

Sam Paech

Building @thinkymachines Maintainer of EQ-Bench https://github.com/orgs/EQ-bench/repositories https://github.com/sam-paech

@sam_paech 4,373 X followersFirst product launched 2y ago

Total monthly visits
345.7K
Down14.9%MoMOct 2026Thinking Machines Lab accounts for 83%
Public revenue
Undisclosed
No public revenue across products
Products
2
Since Sep 2024
Latest launch
14mo ago
Thinking Machines Lab287.3K visits/mo
100% full-depth creator dossier open: complete product portfolio, 12-month combined traffic & MRR curves, shipping cadence, and timeline events.

Traffic

Portfolio · Last 12 months
Monthly visits

Products

Products and research metrics
ProductVisits · OctActions
Thinking Machines LabThinking Machines NEWOne Year of Tinker: Turning Your Expertise into Custom Models287.3K
EQ-Bench 4 Leaderboard我们训练的专项写作模型在 Creative Writing Benchmark 创意写作榜单上获得了第一名,超越了所有的开源模型和闭源模型。 战绩可查:58.4K

Timeline

Launches, marketing, revenue milestones, and pricing/plan/positioning changes across products
  1. Nov 13 · 2d ago

    • Thinking Machines Lab Arena reports open-weights Inkling 2 Small ranks #1 US open model in Code Arena: WebDevReferrals & Backlinks

      Arena reported that Inkling 2 Small by Thinking Machines is now open weights and ranks as the #1 US open model in Code Arena: WebDev at 1647 points.

    • Thinking Machines Lab Ollama adds Inkling 2 Small from Thinking MachinesProduct Activity#Launch Post

      Ollama announced that Inkling 2 Small from Thinking Machines is available on Ollama Cloud with zero data retention. The 594B mixture-of-experts model features a 1M context window.

    • Thinking Machines Lab Arena ranks Inkling 2 Small as #1 US open model in Agent ArenaReferrals & Backlinks

      Arena announced that following its open weights release, Inkling 2 Small by Thinking Machines ranks as the #1 US open model in Agent Arena and #2 among all open models. The model recorded a live +2.9% net improvement score across 11.5k real-world agentic sessions, ranking #22 overall.

    • Thinking Machines Lab details end-to-end 4-bit training of Inkling 2 SmallProduct Activity

      Thinking Machines Lab published details on training Inkling 2 Small using a native 4-bit NVFP4 forward pass in expert layers from pretraining through reinforcement learning on Nvidia GB300 systems. The technique yielded an 8% wall-clock speedup in pretraining and a 20% speedup in RL over full MXFP8 training.

  2. Nov 12 · 3d ago

    • Thinking Machines Lab DesignArena ranks open-weight Inkling 2 Small #1 among open-weight modelsReferrals & Backlinks

      DesignArena reported that Inkling 2 Small by Thinking Machines is now open-weight and ranked #1 among open-weight models on Overall HTML with an Elo of 1384, up 8 Elo points since its initial release. It holds a 32-point lead over Kimi K3 and remains the only American model ranked #1 among open-weight models on DesignArena.