Let's run some numbers on $POD, and address why the FDV is truly a 'meme' in this case: At its last datagen peak the network processed 3.5B output tokens/day across a 600-GPU fleet. With V2 unlocking meaningful per-GPU efficiency gains, 10B output tokens/day is a reasonable conservative target to anchor against. Below is what both look like at OpenRouter-rate pricing ($0.80 per 1M output, $0.15 per 1M input) and the historical 20:1 input:output ratio that real agent workloads tend to land on. Note: this is a single-model exercise (Qwen 3.6 35B economics). In practice the network will serve a basket including Gemma 31B and 26B, which are faster and cheaper, broadening the addressable demand pool. So the math here is conservative on the revenue side. Current 3.5B output/day: Output revenue: $2,800/day Input revenue: $10,500/day Gross revenue: $13,300/day, all of which routes to POD buybacks Node emissions ($0.50 per 1M output): $1,750/day Buyback-to-emissions ratio: 7.6x (!) For every $1 of POD paid out as emissions, $7.60 of POD is being bought back off the market. Most of those emissions are paid bonded, with a 20% fee on liquid claims routing back to stakers, so even the share that does hit the float gets partially absorbed by the same flywheel. Against ~$10M in circulating market cap, that means current-rate buybacks already represent roughly 4% of the float being bought back annually. Post-V2 scenario, closer to 13%. On the float that's actually tradeable, the bid is structurally large. Annualised current: ~$4.85M gross to buyback against a $100M FDV. About 21x. Post-V2 conservative 10B output/day, same prices and mix: Gross revenue: $38,000/day → ~$13.9M/year to buyback Node emissions: $5,000/day Buyback-to-emissions ratio: 7.6x (constant, the design scales linearly) For benchmark, HYPE currently generates around $2.54M/day in fee-driven buybacks, $927M annualised, against a $58B FDV. The market values that buyback flow at roughly 63x. Apply the same multiple to POD's current 3.5B-per-day economics and the implied FDV is roughly $300M, about 3x current. Apply it to the post-V2 10B scenario and you land closer to $880M, around 8.8x current. The 10B figure is a working assumption of what V2 should unlock; the real number could be lower or higher and needs to be seen once V2 ships and the API is live. But the 7.6x buyback-to-emissions ratio is what makes the FDV stop mattering the way most people assume it does. It is not a token waiting to dump on you. It is a token the protocol is mechanically buying back faster than it is being issued.

DeepSeek-V4.1-Flash-1017 is now live on the Dolphin API $10 in free API credits for the next 5,000 signups Lowest pricing on the market at $0.01 / $0.07 per million tokens dphn.ai/API

Running @dphnAI API on my local machine using DeepSeek harness. 99% Cache hit, 0.2$ cost to use 16M tokens. 😍 base:0xed664536023d8e4b1640c394777d34abaff1df8f

The Dolphin Inference API is now live Powered by DeepSeek V4.1 Flash, it delivers Opus 5-class coding & agentic performance at 97% lower cost than frontier models $30 free credits for the first 1,000 signups Start building with the cheapest API on the market → dphn.ai/API