# AI Price-Performance **Entity class:** Economic metric **Domain:** Artificial intelligence / science / political economy **Maturity:** Developed ## Definition **AI Price-Performance** measures how much useful machine capability can be purchased for a given amount of money, compute, energy, or time. ## Mechanism and Consequence Rapid improvement expands the feasible market for intelligence: tasks that were uneconomic become ordinary, while frontier users reinvest savings into larger workloads and create new demand. ## Relationships - [[wiki/AI Economy|AI Economy]] - [[wiki/Abundance Economics|Abundance Economics]] - [[wiki/Compute Scarcity|Compute Scarcity]] - [[wiki/AI Progress|AI Progress]] ## Source Dossiers - [[research/AI Is Now So Good They Will Pay You to Quit|AI Is Now So Good They Will Pay You to Quit]] - [[research/Super Intelligence Week Index of Statements from the White House Luncheon|Super Intelligence Week Index of Statements from the White House Luncheon]] - [[research/Jensen Pushes Back on Doomers Moonshots Live|Jensen Pushes Back on Doomers Moonshots Live]] ## Simple Reminders, Quotations, and Thoughts > The abundance thesis is visible in AI price-performance: the models are becoming twice as capable and twice as cheap at an accelerating rate. > **— Adapted from Peter Diamandis**, *MOONSHOTS Live, September 25, 2026* [[reminders/Abundance/AI Is Becoming Twice as Capable and Twice as Cheap by Peter Diamandis|AI Is Becoming Twice as Capable and Twice as Cheap by Peter Diamandis]] > "The intelligence reaches a certain point where getting smarter isn’t necessarily the most impactful thing. Getting cheaper becomes the most impactful thing." > **— Tom Bilyeu**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/AI Access/Cheaper Intelligence Can Matter More Than Smarter Intelligence by Tom Bilyeu|Cheaper Intelligence Can Matter More Than Smarter Intelligence by Tom Bilyeu]] > The prediction that agentic AI will transform the economy within two years depends upon the cost falling dramatically. At today’s prices, deploying enough high-end agents to produce that transformation would be too expensive. > **— Adapted from Tom Bilyeu**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/AI Access/Agentic Economic Transformation Depends on Falling Costs by Tom Bilyeu|Agentic Economic Transformation Depends on Falling Costs by Tom Bilyeu]] > The cost of AI is coming down, and the decline in cost changes which agentic systems can be deployed economically. > **— Adapted from Emad Mostaque**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/AI Access/Falling AI Costs Decide Which Agentic Systems Become Economic by Emad Mostaque|Falling AI Costs Decide Which Agentic Systems Become Economic by Emad Mostaque]] > "You can build a website as good as Fable with this incredibly cheap model if you architect and harness it correctly — for 100 times cheaper." > **— Emad Mostaque**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/AI Access/AI Architecture Can Cut Software Costs One Hundredfold by Emad Mostaque|AI Architecture Can Cut Software Costs One Hundredfold by Emad Mostaque]] > "When o3, the first reasoning model, came out, it was the first model to get the gold medal in the International Math Olympiad. To do that cost $80,000. Today you can do it for $10 with the latest model." > **— Emad Mostaque**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/AI Access/Olympiad-Level Reasoning Fell From Eighty Thousand Dollars to Ten by Emad Mostaque|Olympiad-Level Reasoning Fell From Eighty Thousand Dollars to Ten by Emad Mostaque]] > "With the additional chips, with the harnesses and others, you should expect a 100-times decrease [in price] for an equivalent performance between now and one year from now." > **— Emad Mostaque**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/AI Access/Equivalent AI Performance Could Become One Hundred Times Cheaper in a Year by Emad Mostaque|Equivalent AI Performance Could Become One Hundred Times Cheaper in a Year by Emad Mostaque]] > "The costs are dropping, but then the next generation of chips are coming out. Vera Rubin, Nvidia’s new chip, is a ten-times improvement in cost for inference." > **— Emad Mostaque**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/AI Infrastructure/New Chips Can Improve Inference Cost Tenfold by Emad Mostaque|New Chips Can Improve Inference Cost Tenfold by Emad Mostaque]] > "While you need frontier models to be close to capability, you don’t want [to etch them]. The moment you have a good-enough model, then you do want that, because it costs a fraction of the price and it’s 100 times faster, a hundred times cheaper. When you use a model that goes at 1,000 words a second versus 50, it’s night and day. [AI output] is instantaneous, and you’re like: well, I’m the slow one. The latest MiniMax H3 model is eight seconds now to generate a 15-second clip. As a human, I need 15 seconds to see this, and then I’ll need more time." > **— Emad Mostaque**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/AI Access/Good-Enough Models Can Run One Hundred Times Faster and Cheaper by Emad Mostaque|Good-Enough Models Can Run One Hundred Times Faster and Cheaper by Emad Mostaque]] > "There is a model from the Chinese [company] Alibaba called Qwen 27B. It’s a 27-billion-parameter model. … It’s like 60 GB. You need 60 GB of RAM to run it by default. They got it down to like 25. A team called PrismML got it down to six. … Same model." > **— Emad Mostaque**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/Local AI/The Same AI Model Shrunk From Sixty Gigabytes to Six by Emad Mostaque|The Same AI Model Shrunk From Sixty Gigabytes to Six by Emad Mostaque]] > The compressed Qwen 27B model matches the frontier Opus model that you were using in January, and it runs on 6 GB — which is a Raspberry Pi, which uses as much energy as your brain. > **— Adapted from Emad Mostaque**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/Local AI/Frontier AI Now Runs on the Energy Budget of a Human Brain by Emad Mostaque|Frontier AI Now Runs on the Energy Budget of a Human Brain by Emad Mostaque]] > "Five years from now will be the biggest infrastructure boom we’ve ever seen in our lives, both from government programs to stimulate it — because it’s the easy thing for them to stimulate — [and] because it’s kind of needed. The infrastructure is a bit creaking." > **— Emad Mostaque**, *Impact Theory with Tom Bilyeu, 2026* [[reminders/AI Infrastructure/AI and Robotics Could Produce the Largest Infrastructure Boom of Our Lives by Emad Mostaque|AI and Robotics Could Produce the Largest Infrastructure Boom of Our Lives by Emad Mostaque]] > "Claude Opus 5.5 delivers performance equivalent to Fable 5.1 at literally half the price." > **— Peter Diamandis**, *MOONSHOTS Live, September 25, 2026* [[reminders/AI Access/Frontier AI Performance Fell to Half the Price by Peter Diamandis|Frontier AI Performance Fell to Half the Price by Peter Diamandis]] > "A year or two ago, one school of thought held that hallucinations would be a disaster and never go away. Like everything else, we have scaled out of it, and the error rate is plummeting through pure scale." > **— Dave Blundin**, *MOONSHOTS Live, September 25, 2026* [[reminders/AI Access/AI Hallucination Rates Are Plummeting Through Scale by Dave Blundin|AI Hallucination Rates Are Plummeting Through Scale by Dave Blundin]] > "Across 118 real app builds, Claude Sonnet 5.5 produced apps that scored level with Opus 5. It got there in 3.6 iterations per build on average, where Opus 5 took 7.7. It had the fewest failed tool calls of any model we compared." > **— Gabriel Grinberg**, *Introducing Claude Sonnet 5.5, September 28, 2026* [[reminders/AI Access/A Lower-Cost AI Matched a Frontier Model in Half as Many Iterations by Gabriel Grinberg|A Lower-Cost AI Matched a Frontier Model in Half as Many Iterations by Gabriel Grinberg]] > "On our private suite of 2,441 finance tasks covering Q&A, extraction, analysis, and forecasting, Claude Sonnet 5.5 scored ahead of Sonnet 5 and used about 121k tokens per answer, where Sonnet 5 used 497k. On our analyst search and retrieval work, it was better than Sonnet 5 in almost every way." > **— Joe Poirier**, *Introducing Claude Sonnet 5.5, September 28, 2026* [[reminders/AI Access/AI Cut Financial-Analysis Token Use by More Than Seventy-Five Percent by Joe Poirier|AI Cut Financial-Analysis Token Use by More Than Seventy-Five Percent by Joe Poirier]] > "With millions of Rovo-assisted actions powering our customers' workflows each month, execution speed is critical. Claude Sonnet 5.5 will allow teams to run their Rovo Agents up to 30% faster than they could with Sonnet 5." > **— Jamil Valliani**, *Introducing Claude Sonnet 5.5, September 28, 2026* [[reminders/Agentic AI/Enterprise AI Agents Can Now Run Thirty Percent Faster by Jamil Valliani|Enterprise AI Agents Can Now Run Thirty Percent Faster by Jamil Valliani]] > "NVIDIA accelerated computing delivers value across generations. CoreWeave's NVIDIA V100 GPUs are still running customer workloads nearly a decade after Volta launched, even as CoreWeave brings Vera Rubin NVL72 into production. That's the strength of the NVIDIA platform: infrastructure that keeps earning for years, and the flexibility to put the right GPU on the right workload." > **— Ian Buck**, *From Training to Production, NVIDIA and CoreWeave Close the Loop on Agentic AI, September 30, 2026* [[reminders/AI Infrastructure/AI Infrastructure Can Keep Earning Across Hardware Generations by Ian Buck|AI Infrastructure Can Keep Earning Across Hardware Generations by Ian Buck]] > "CoreWeave's deployment of Vera puts 128 CPUs and 11,264 cores in a single rack, enough for more than 11,000 concurrent environments at one core each." > **— NVIDIA**, *From Training to Production, NVIDIA and CoreWeave Close the Loop on Agentic AI, September 30, 2026* [[reminders/AI Infrastructure/One AI Rack Can Run More Than Eleven Thousand Concurrent Agent Environments by NVIDIA|One AI Rack Can Run More Than Eleven Thousand Concurrent Agent Environments by NVIDIA]]