Mira Murati, former CTO of OpenAI, launched her first AI model through Thinking Machines Lab after a two-year absence from public view. The model, named Inkling, now runs on OpenRouter and targets the open-source AI space where competition from Meta's Llama, Mistral, and others has intensified.
Inkling's MCP (Model Context Protocol) score shows legitimately strong performance metrics. The model demonstrates competitive reasoning capabilities and language understanding that places it among the leading open-source options available in the Western market. Early benchmarks suggest Murati's team focused on efficient scaling rather than pure parameter count.
Pricing presents the real complexity. Inkling undercuts some competitors on raw token costs, but total cost-of-ownership calculations factor in inference speed, throughput, and error rates. Users evaluating deployment at scale must weigh token price against actual latency and accuracy requirements. For budget-conscious developers, the math works. For enterprises prioritizing speed and reliability, more expensive alternatives may offer better economics.
The launch marks Murati's first major project since departing OpenAI in June 2023. Her credibility and the lab's technical foundation generated immediate attention from the open-source community. OpenRouter integration ensures easy access and streamlined billing for users already familiar with that platform.
Inkling positions itself between Meta's aggressive open-source strategy and proprietary models like GPT-4. The model targets developers who want state-of-the-art performance without vendor lock-in, though it lacks the training data advantages and compute resources that OpenAI commands.
Adoption will depend on real-world performance data from production environments. Benchmark scores matter less than consistency across diverse use cases. Early interest suggests the model fills a gap, but sustained traction requires proof that price-to-performance claims hold up under commercial workloads. The next months will determine whether Inkling becomes a standard reference point or remains a specialized option for cost-conscious teams.