Realtime AI News
Tokens Infinity AI Raises Hundreds of Millions in Three Rounds in One Year: Ex-ByteDance AI Coding Lead and Tsinghua Yao Class CTO Build Enterprise AI Agent Infrastructure
Tokens Infinity AI (词元无限), co-founded by former ByteDance MarsCode and Trae lead Yang Ping and Tsinghua Yao Class graduate Wang Wei, has completed three funding rounds totaling hundreds of millions of yuan in its first year. The latest round was led by Linxin Investment with follow-on from Huakong Fund.
Tokens Infinity AI, an enterprise AI Agent infrastructure company, has completed three funding rounds totaling hundreds of millions of yuan in just one year since its founding. The latest round was led by Linxin Investment, with existing investor Huakong Fund also participating.
CEO Yang Ping previously built ByteDance's AI coding products MarsCode and Trae, leading the AI transformation of a R&D organization with tens of thousands of engineers. CTO Wang Wei, a graduate of Tsinghua University's prestigious Yao Class, served as a chief architect at a Fortune 500 company and later at a robotics unicorn.
The company positions itself not as another AI coding tool vendor but as an AI Agent infrastructure provider for enterprise software engineering. Yang Ping argues that the real opportunity in AI coding lies not in consumer tools but in enterprise R&D environments where what matters is shipping velocity and quality — not lines of code generated.
Tokens Infinity's product stack follows a pyramid structure. At the base are domain-specific agents: InfCode (digital programmer) and InfTest (digital test engineer), connected through a 'generate-test-verify' loop that turns probabilistic LLM outputs into reliable deliveries.
The middle layer handles engineering understanding and knowledge capture: DeepMap builds code knowledge graphs for complex legacy systems, HarnessAgent encodes enterprise processes into agent guardrails, and CodeReview manages code quality. At the top, TokenHub handles model routing and cost optimization, reducing token costs by over 25%.
The company's most distinctive feature is its business model: transitioning from seat-based or API-call pricing to 'Result-as-a-Service.' Yang Ping says enterprise customers pay for outcomes, not tools. The company targets 2x to 4x team-level efficiency gains rather than the more common '10x for every developer' claims.
In an increasingly crowded AI coding market, Tokens Infinity has chosen a harder path — going deep into enterprise R&D sites, dealing with messy legacy systems, rigorous testing requirements, and security boundaries. The three funding rounds in rapid succession validate this bet in a challenging fundraising environment.
Why it matters
Tokens Infinity AI's rapid fundraising cadence and 'Result-as-a-Service' model signal that the AI coding market is evolving from individual productivity tools to enterprise-grade software engineering infrastructure.
Nearby Updates
All07/30, 14:14
Infinigence Releases PDD Cross-Cluster Inference Architecture: First-Token Latency Down 51.5%, Cost Down 37.5%
Infinigence has released the full technical report for its PDD cross-cluster heterogeneous inference architecture, first disclosed at WAIC 2026. The architecture deconstructs the traditional PD separation link into three layers, achieving 51.5% reduction in first-token latency and 37.5% cost reduction per token on an 80Gbps WAN Ethernet environment.
07/30, 13:55
Anthropic Reaches $1.5 Billion Data Copyright Settlement, Raising Questions on AI Training Data Boundaries
Anthropic has agreed to a roughly $1.5 billion settlement in a data copyright lawsuit, one of the largest such deals in the AI industry. The case centered on the legality and boundaries of AI training data sourcing, setting a significant precedent for data compliance practices across the sector.
07/30, 13:51
KPMG Canada named OpenAI Elite Partner to accelerate enterprise AI deployment
KPMG Canada has formed a strategic collaboration with OpenAI, earning the top-tier Elite Partner designation in the OpenAI Partner Network. The alliance combines OpenAI's technology with KPMG's business transformation, governance, and risk management expertise to help Canadian organizations scale AI from experimentation to enterprise-wide deployment.
07/30, 13:48
Google DeepMind Disbands Nobel-Winning AlphaFold Team, Shifts Resources to Gemini
Google DeepMind has disbanded its AlphaFold team, the Nobel Prize-winning protein structure prediction group, and reallocated resources to the Gemini project. The restructuring signals a major strategic shift from pure scientific research toward commercial AI product development.