Thinking Machines releases Inkling, a multimodal AI model with open weights for efficient reasoning across text, image, and audio.

{
  "signal_id": "8bb8bf7a-4b84-4663-aa2e-93abeed040dc",
  "signal_type": "linkedin-post-company",
  "signal_subtype": "linkedinPost",
  "detected_at": "2026-07-22T13:25:14.39+00:00",
  "company": {
    "name": "Thinking Machines Lab",
    "domain": "thinkingmachines.ai"
  },
  "data": {
    "tags": [
      "Artificial Intelligence",
      "Product Launch",
      "Open Source",
      "Data Science"
    ],
    "summary": "Thinking Machines releases Inkling, a multimodal AI model with open weights for efficient reasoning across text, image, and audio.",
    "post_url": "https://www.linkedin.com/posts/thinkingmachinesai_inkling-our-open-weights-model-activity-7483224333616590848-Oys1",
    "num_likes": 1851,
    "post_text": "Today, we are introducing Inkling.\n\nInkling reasons efficiently across text, image, and audio modalities. We are making the full weights available.\n\nhttps://lnkd.in/gY4NvS5h\n\nAvailable today for fine-tuning on Tinker. Play with it in the Inkling Playground.\n\nCost and latency are important in real-world use cases. Inkling's continuous thinking effort lets you pick your point on the cost/performance curve - reaching the same score with a fraction of the tokens.\n\nInkling natively understands and reasons across text, audio, and images. It’s strong on audio in particular, ranking among the strongest open-weights models on VoiceBench, MMAU, and AudioMC.\n\nWe’re grateful to our partners for their day-0 support across the open-source ecosystem: NVIDIA, Together AI, Fireworks AI, Databricks, Unsloth AI, Modal, Baseten, LightSeek Foundation, Inferact on VLLM, RadixArk on SGLang.\n\nInkling is the first in a family. We’ve included some details of Inkling-Small, a lighter-weight model trained on a similar recipe, with full weights to follow.\n\nWe hope you enjoy Inkling, and as always we’re keen to see what you build.",
    "initiatives": [
      {
        "topic": "release multimodal AI model Inkling",
        "urgency": 0.9
      },
      {
        "topic": "release Inkling-Small model",
        "urgency": 0.7
      }
    ],
    "posted_date": "2026-07-15T18:21:52.719Z",
    "num_comments": 41,
    "technologies_mentioned": [
      {
        "name": "Inkling",
        "status": "implemented"
      },
      {
        "name": "Tinker",
        "status": "using"
      },
      {
        "name": "NVIDIA",
        "status": "partnering"
      },
      {
        "name": "Together AI",
        "status": "partnering"
      },
      {
        "name": "Fireworks AI",
        "status": "partnering"
      },
      {
        "name": "Databricks",
        "status": "partnering"
      },
      {
        "name": "Unsloth AI",
        "status": "partnering"
      },
      {
        "name": "Modal",
        "status": "partnering"
      }
    ]
  }
}

curl https://signals.autobound.ai/v1/signals/8bb8bf7a-4b84-4663-aa2e-93abeed040dc \
  -H "X-API-KEY: $AUTOBOUND_API_KEY"

curl -X POST https://signals.autobound.ai/v1/companies/enrich \
  -H "X-API-KEY: $AUTOBOUND_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"domain":"thinkingmachines.ai","limit":20}'