Artificial intelligence inference routing startup OpenRouter Inc. today announced it has raised $113 million in new funding led by CapitalG, Alphabet Inc.’s independent growth fund, to push the envelope for providing access to generative AI models.

Additional investors participating in the Series B round included Nvidia Corp.’s venture capital arm NVentures, ServiceNow Ventures, MongoDB Ventures, Snowflake Ventures and Databricks Ventures, alongside existing investors Andreessen Horowitz and Menlo Ventures.

OpenRouter provides a quick and easy interface for individual developers and enterprise businesses to access, route to and optimize for centralized closed-source AI model providers and open-source models. It has a marketplace of hundreds of models and providers, allowing AI apps and agents to pick and choose the best possible model and provider through a single unified interface.

At the same time, the company’s interface provides a single pane for billing, tracking usage and inference and enforcing rules. Organizations can add policies such as per-request data handling policies, team-level access and routing permissions, providing a way to cap spending, gain spend-level visibility and support audit-friendly usage reporting, alongside intelligent routing to improve cost and performance.

For uptime-conscious businesses, the company also provides intelligent failover to increase reliability. Having access to multiple vendors through a single, easy-to-use interface means that developers and businesses can experiment without vendor lock-in or risk.

“Running inference at scale is fundamentally a multimodel problem,” said co-founder and Chief Executive Alex Atallah. “The era of picking a single model is over. Success now depends on continuously routing across a changing market.”

The rising adoption of AI and the associated cost

It’s not uncommon for a company to consume almost a billion tokens a month in inference. A token is the atomic unit of data processing used by language models such as ChatGPT. It is often a word, part of a word or a punctuation mark. Tokens cost money to process in both input and output.

For an example of scale, OpenAI Group PBC’s GPT-5.5 inference costs $5 per 1 million input tokens and $30 per 1 million output tokens, double the price of 5.4. That is not unexpected, given GPT-5.5 is a powerful flagship frontier model provided by a centralized industry leader. OpenRouter also gives access to smaller open-source models that are significantly less expensive but still powerful in their own right, such as Qwen3.7 Max at $2.50 per 1 million input tokens and $7.50 per 1 million output tokens.

According to a recent Deloitte Touche Tohmatsu Ltd. study, 67% of enterprise businesses already compute almost 1 billion tokens a month. As more AI agents push into the marketplace and users adjust to models that can reason over audio, images and video, the number of tokens being consumed will only balloon further.

At the same time, companies are looking for ways to contain costs by routing different types of needs between models optimized for specific kinds of reasoning. For example, a company might send summarization tasks to a smaller text-only model, while having a multimodal experience, such as audio or video, handled by a more costly model designed specifically for that purpose. Its answers could then be passed back to a smaller orchestrator model that can parse the output alongside business information.

These competing cost and performance needs create a level of operational complexity that means organizations must either build their own tooling to support this sort of routing, or sign up for a platform such as OpenRouter that provides a centralized control plane to optimize cost, latency and capability.

OpenRouter provides its own usage and model rankings as open data, which have become widely referenced signals across the AI industry for trends in real-world adoption of particular models, performance dynamics and pricing decisions. The company’s volume has surged to about 25 trillion tokens per week, representing a fivefold increase from 5 trillion tokens per week six months ago. The platform currently supports more than 8 million global users, including AI-native startups and large enterprises.

The company said it intends to use the new funding to expand its routing, governance and optimization capabilities as enterprises continue to adopt, deploy and scale AI throughout their operations.

Image: SiliconANGLE/Microsoft Designer