Together AI and Pearl Research Labs Team Up to Reduce the Cost of AI Inference
Quick Answer
Together AI partners with Pearl Research Labs to launch a discounted inference endpoint for the Gemma-4-31B-it-pearl model, leveraging Pearl's blockchain protocol to reduce AI workload costs by over 25% through cryptocurrency emissions.
Quick Take
Together AI partners with Pearl Research Labs to launch a discounted inference endpoint for the Gemma-4-31B-it-pearl model, leveraging Pearl's blockchain protocol to reduce AI workload costs by over 25% through cryptocurrency emissions. This collaboration aims to enhance GPU economic throughput and lower LLM price-per-token.
Key Points
- New inference endpoint for Gemma-4-31B-it-pearl offers over 25% discount.
- Pearl Network's Proof of Useful Work generates cryptocurrency to offset AI costs.
- Pearl's technology allows GPUs to produce digital assets without extra costs.
- Future plans include expanding Pearl-powered product offerings and currency emissions.
- Partnership aims to change AI unit economics and improve GPU efficiency.
📖 Reader Mode
~2 min readTogether AI’s mission is to make AI accessible and widely distributed. Today, we are excited to announce an exclusive partnership with Pearl Research Labs, the creators of the Pearl Network, a blockchain protocol. This protocol uses Proof of Useful Work to harness the computation of inference and training workloads to secure a blockchain and mine a new cryptocurrency that allows offsetting the cost of AI workloads.
Pearl’s mathematical breakthrough is to generate the Pearl cryptocurrency by obtaining strong proofs of matrix multiplications that happen in the course of forward and backward passes of training and inference computations. In collaboration with Pearl Research Labs, we're introducing a new inference endpoint for Gemma-4-31B-it-pearl, an instruction-tuned model. This endpoint is discounted by over 25%, offset by the future value of crypto emissions. As the spot price of ¶PRL increases, we expect to use more of the value generated by Pearl emissions to further reduce the price of this endpoint.
This is our first product that integrates the Pearl Network. Eventually, we plan to expand our Pearl powered product portfolio, and offer ways for our customers to directly obtain a portion of the currency emissions.
Pearl changes the unit economics of AI, by allowing every GPU cycle powering AI training and inference to simultaneously produce a native proof-of-work digital asset—the ¶PRL coin—at no additional cost. As such, Pearl’s breakthrough technology increases the economic throughput of GPUs and decreases LLM price-per-token.
Pearl Research Labs is excited to partner with Together AI and its world-class research and inference engineering teams to deliver the first inference endpoint that is discounted on account of the Pearl coins being generate in parallel to inference tokens.
– Omri Weinstein, Co-founder and CEO, Pearl Research Labs
Learn more about the protocol at Pearl Research Labs and try out our first Pearl endpoint at www.together.ai/models/gemma-4-31b-it-pearl!
— Originally published at together.ai
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from Together AI
See more →Serving MiniMax-M3 for efficient inference: Unlocking 1M-Token Context and Multimodality Without Regrets
MiniMax's M3 model introduces a 1M-token context and multimodal capabilities, optimized for efficient inference with a 9x speedup in prefill and 15x in decoding, supported by Together AI's cloud infrastructure.
