if you have a .edu, .gov or .mil email you can now get a full 8x B300 node for $4/GPU-hour. nebius charges $9.50 on demand and a lot of clouds are sold out AMD MI355X nodes are $3 and the wait right now is under a minute you just ask marshall, their AI agent, and it finds you a node this is a major public good for researchers who've been priced out of frontier compute thank you @AnjneyMidha

anthropic and openai have completely opposite compute strategies anthropic buys from everyone. $111b with google for TPUs, $110b with amazon for trainium, $31b with microsoft, and ~$135b with neoclouds like spacex, nscale, lambda, and volta openai bet on doing compute itself. $300b with oracle for stargate, $250b with microsoft, and its own stargate campuses in texas, ohio, and georgia anthropic signed roughly 3x more neocloud capacity than openai, and around two thirds of its commitments are on non-nvidia chips

microsoft just pulled out of nscale’s uk data center deal. which neocloud is next? here’s every major one by contracts, revenue, debt, borrowing rate, and how much of their capacity is actually live

most of what you pay for a guaranteed h100 is the guarantee -> hyperscaler on demand ~$7.15/hr -> neocloud average ~$2.83/hr -> interruptible on vast from $0.53/hr an idle gpu costs its owner like $0.10/hr in power so basically any price beats leaving it unused. electricity markets already work this way, fixed rates for stuff that can’t wait, real time prices for everything else, and big users paid to shut off when the grid is tight. @sfcompute letting you resell reserved compute you’re not using is the same logic inference is the workload that can buy compute like this since batch jobs, evals, and background agents can wait. you already see it in tokens, batch is half off at openai and anthropic and on auction markets like liquid inference from @Architect_Fi, open models clear way below the model maker’s price

a v100 from 2017 still rents for about twice its resale value every year -> a used 32gb card sells for ~$600-700 -> it rents for about $0.15/hr on vast, ~$1,300 for a full year worth remembering when people say gpus are worthless after 3 years. that's a big part of the neocloud/compute bear case right now

open model tokens are getting close to free on liquid inference -> deepseek v4.1 flash is $0.001 in / $0.008 out per million tokens. deepseek charges $0.15 / $0.60 -> kimi k3 is $0.02 / $0.70 vs $3 / $15 from moonshot -> glm 5.3 is $0.015 / $0.21 vs $1.40 / $4.40 from zhipu once enough sellers compete on the same model, tokens behave like any other commodity. it's much easier to list here than on openrouter, so people with idle gpus just jump in and undercut to fill them the market design is doing the work. live quotes and an auction on every request let spare capacity clear near cost, which a fixed rate card can't do video just launched with one seller per model at fal's list price. looking forward to watching that market too

modal is a good example of where compute markets are heading. they sell gpus by the second with no commitment for less than the big neoclouds charge on demand -> b200 is $6.25/hr on modal vs $8.50 at nebius -> b300 is $7.10 vs $9.50 -> h200 is $4.54 vs $5.40 they rent from 20+ clouds and commit to capacity ahead of demand. pooling demand across all their customers keeps utilization high, which is what lets them price lower the more compute trades like a commodity, the more value goes to whoever aggregates supply and matches it to demand

you can get kimi k3 on liquid inference for $0.39 per million input tokens and $0.57 output right now moonshot's own api is $3 and $15. the cheapest provider on openrouter is $0.73 and $10 it's still cheaper after their cache discounts too the difference is market design. on openrouter each provider lists a fixed price and openrouter spreads traffic across them. on liquid every request is an auction, providers can reprice any minute, and the cheapest offer that meets your speed and quality rules wins markets will exist for compute and tokens just like for other commodities

the token index drop is mostly an explosion in cheap model usage, not less frontier spend on openrouter models under $0.25 per million tokens are now 3/4 of all tokens, and that shift alone cut the average price per token ~31% over the last month frontier usage keeps growing. tokens on the top closed models grew 55% over the last month and they're still about half of all spend and there will always be demand for frontier tokens. for r&d and anything adversarial like negotiation or trading, you want the best model meanwhile b200 rental prices are up 67% since mid june

google knew the risk of cutting noam shazeer's compute (the stock fell by $100 billion the day his departure was announced) but did it anyway because they didn't have enough dario personally calls ppl to beg for compute imo ppl are still underestimating the importance and scarcity of compute it's even harder for startups who can't fund long multiyear contracts or put 30% down compute is the oxygen of ai

agree, open tokens burn about as much compute as closed tokens. inference providers (like @baseten and @FireworksAI_HQ) have positive margins on top of their compute costs, even with dozens of companies competing to serve the same models on openrouter there’s a question of why. in a world where AI can spin up an efficient inference stack for anyone, a large software edge shouldn’t last long a lot of it is that inference providers signed long-term compute contracts making them effectively long compute. those contracts have gotten more valuable since they signed. a 1-year H100 contract went from ~$1.70 last october to ~$2.75 today, and B200 rentals are up ~85% since june so part of that margin is effectively a compute trade rather than purely a software moat

a GPU cluster can pass every benchmark and still go down for days IREN’s response was NVIDIA exemplar status, which shows its cluster hits NVIDIA’s reference performance on their benchmark recipes. whether a site stays up month after month is a separate question here’s what I’d ask before signing a GPU contract - which site am I on, and is power, cooling and network N+1 or N+0 - does the provider run that datacenter or rent colo space and rely on the landlord’s techs - what was uptime at that site over the last 90 days, and how does the SLA define downtime - how fast do health checks catch a bad GPU and pull it out of your jobs, and are there hot spares - on NVL72, what happens when a tray dies. it takes 4 GPUs out of the NVLink domain, so ask if they run the rack degraded or swap it and how fast - which XIDs they act on. the ones that mean bad hardware are 79 (GPU fell off the bus), 48, 94 and 95 (memory errors) and NVLink errors (74 on hopper, 144 to 150 on blackwell). 13, 31 and 43 are usually your own code - was the cluster burned in with the GPUs and network stressed at the same time - link flaps and error counters on the infiniband or RoCE network the providers with good answers usually love being asked

some inference providers on openrouter are playing a funny game openrouter's routing seems to reward whoever lists the cheapest input price so a few providers price input near zero and make it back on output on deepseek v4 flash today, relace has the lowest input price ($0.01 per million tokens) and gets 33% of all traffic. but it charges $1.28 per million output tokens while streamlake charges $0.08 even though 94% of tokens on this model are input, relace still ends up almost 4x more expensive than streamlake. for reasoning-heavy work it's about 15x in theory openrouter should be a really good marketplace where competition drives prices down fast. but that only works if the incentives reward the cheapest real price, not gaming the listings. it's a really important market design problem

the big AI clouds are building in really different ways right now Google is unique in that it owns basically every layer. it designs its own TPU chips, builds and runs its own data center campuses, and owns the application layer with Gemini, Search, YouTube, etc. its cloud business runs a very healthy operating margin. SpaceXAI is similarly very integrated. they've put their own gas turbines and Tesla Megapacks next to their data centers and run Grok on a lot of their compute, though they still buy every GPU from Nvidia. Crusoe develops the sites and the power, although for its biggest sites like Abilene, Blue Owl owns most of the campus. Nebius owns its Finland site, designs its own servers, and fun fact, pipes the server heat into the local town! CoreWeave takes a different approach and leases every data center, only owning the GPUs and software. Fireworks goes further and mostly rents GPUs from a bunch of providers to sell inference, though we're seeing more inference providers start buying their own compute to expand margins and bring on more capacity the more of the stack you own, the more capital it takes upfront, but also the better you can work around bottlenecks, which are getting more pervasive as the compute shortage worsens