Thinking Machines partners with Bridgewater to build AI model that cuts errors by nearly 30%

{
  "signal_id": "678ba7d8-dc73-a78c-eae6-4c9e65abe17f",
  "signal_type": "news",
  "signal_subtype": "partners_with",
  "detected_at": "2026-07-03T05:16:17+00:00",
  "company": {
    "name": "Thinking Machines Lab",
    "domain": "thinkingmachines.ai"
  },
  "data": {
    "url": "https://cryptobriefing.com/thinking-machines-bridgewater-ai-model-accuracy/",
    "title": "Thinking Machines partners with Bridgewater to build AI model that cuts errors by nearly 30% - Crypto Briefing",
    "excerpt": "Mira Murati's startup and the world's largest hedge fund built a custom model that outperforms GPT, Claude, and Gemini on financial document tasks while slashing inference costs by 13.8x Share When the world’s largest hedge fund decides its analysts are spending too much time on document busywork, it doesn’t just buy a ChatGPT subscription. Bridgewater Associates teamed up with Thinking Machines Lab, the AI startup founded by former OpenAI CTO Mira Murati, to build a custom fine-tuned model that reduces errors by 29.8% compared to the best available frontier models. The results, published June 30 by Bridgewater’s AIA Labs and Thinking Machines Lab, show the specialized model hitting 84.7% average accuracy across six information-filtering tasks. Leading models like GPT, Claude, and Gemini variants, even when juiced with expert prompt engineering, were stuck in the mid-70s. The model was constructed on the Qwen3-235B base and trained using Thinking Machines’ proprietary Tinker platform. Two training techniques did most of the heavy lifting: interleaved batching delivered a 12.1% accuracy boost, while on-policy distillation added another 3.1%. The Tinker API handled the infrastructure side, letting the team iterate rapidly without managing GPU clusters directly. Inference costs dropped by a factor of 13.8x per task compared to frontier models. The six tasks the model handles fall...",
    "summary": "Thinking Machines Lab partnered with Bridgewater Associates to build a custom fine-tuned AI model for financial document tasks, which reduces errors by 29.8% compared to leading models.",
    "planning": false,
    "image_url": "https://static.cryptobriefing.com/wp-content/uploads/2026/07/03011610/billionaire-ray-dalio-is-betting-big-on-these-4-ai-stocks-1-800x420.jpeg",
    "confidence": 0.95,
    "published_at": "2026-07-03T05:16:17Z",
    "article_sentence": "Bridgewater Associates teamed up with Thinking Machines Lab, the AI startup founded by former OpenAI CTO Mira Murati, to build a custom fine-tuned model that reduces errors by 29.8% compared to the best available frontier models.",
    "related_company_name": "Bridgewater Associates",
    "related_company_domain": "bridgewater.com"
  }
}

curl https://signals.autobound.ai/v1/signals/678ba7d8-dc73-a78c-eae6-4c9e65abe17f \
  -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}'