Sapiom, which helps businesses build, ship, and scale AI agents and lower token costs, raised a $35M Series A led by Dragonfly, taking its total funding to $50M
THE SO WHAT
Routing workloads to the lowest-cost tokens is the first step toward a meta-coordination layer that arbitrages across models, providers, and pricing in real time. If you’re building agentic systems at scale, you now have to assume your customers will expect dynamic cost routing rather than static vendor lock-in.
READ THE SOURCE
MORE FROM THE WIRE
Startups & VentureWispr raises $280M at $2B valuation as it looks beyond dictation
A $280M round at a $2B valuation for Wispr says investors are betting that voice is a primary interface for applied AI, not a feature. If your product has frequent, short workflows — field work, sales, support — assume competitors will layer in conversational agents and start testing where voice-first actually reduces friction.
Startups & VentureGravis Robotics raises $200M for autonomous construction
A $200M raise for Gravis Robotics to retrofit excavators and heavy equipment is a bet that autonomy will hit construction before many white-collar sectors feel real AI displacement. Developers and EPCs should start bidding projects assuming mixed human–robot fleets — and lock in partners now while vendor capacity is still scarce.
Startups & VentureGroq Valued at $3.5 Billion in Funding Round After Nvidia Deal
Groq raising $350M at a $3.5B valuation — roughly half its pre-Nvidia-deal mark — shows how hard it is for independent silicon to hold pricing power once it partners with the dominant GPU supplier. For AI infra buyers, this is leverage: expect more aggressive pricing and licensing terms from alternative accelerators trying to stay relevant.
Startups & VentureVoice AI startup Wispr raised a $280M Series B led by Menlo Ventures at a $2B valuation, taking its total funding to $361M, and now has 100K business customers
A $2B valuation and 100K business customers for a voice-first interface company says investors are underwriting voice as a primary UI, not a niche accessibility feature. If your product still treats voice as an add-on, assume your users will soon compare you to workflows that start and end in speech.