iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
According to an Epoch AI report published September 22, 2026, the estimated cost of reaching a given level of AI performance fell about 47% per quarter—roughly 13-fold per year—across five benchmarks since 2023. That is faster than the historical price declines the report measured for compute, DNA sequencing, lithium-ion batteries and electricity. But it is not evidence that every AI subscription or task has become 13 times cheaper: the estimate tracks the lowest-cost model found for a target benchmark score, not a general price index.
What does “the price of AI” mean in this report?
Epoch AI’s report, “The plunging price of thought”, follows the estimated cost of using the cheapest model able to reach specified scores on five benchmarks covering mathematics, hard sciences and games of skill. It tracks costs at fixed performance levels over roughly three years. The figures are therefore a cost frontier: they describe what a user might pay if they searched among available models for a low-cost option that met a particular benchmark target.
This is not the average price of AI products, the cost of completing every real-world workflow, or a measure of AI’s total economic value. Benchmark performance is only a proxy for useful work, and the cheapest model for a benchmark may not be the right choice for a person’s task.
Free tools Windows power users keep installed
One-click scans. No signup required.
How quickly does Epoch AI estimate costs are falling?
Across the five benchmarks, Epoch AI estimates a decline of about 47% per quarter, or about 13-fold per year, since 2023. The pace differs by task: the report estimates quarterly declines of 39–43% for game-based puzzles and 50–52% for math problems.
#1 Best Overall
- Built for Local AI Development: AMD Ryzen AI Halo is designed for local AI development and inference, featuring 128GB unified memory and support for up to 200B parameter models to build and run intensive AI workloads locally.
- 128GB Unified Memory: Features 128GB LPDDR5x unified memory at 8000 MT/s with 256 GB/s memory bandwidth, providing a shared memory pool across the CPU, GPU, and NPU to support larger AI models.
- AMD Ryzen AI Max+ 395 Processor: Features 16 cores, 32 threads, and Zen 5 architecture, paired with AMD Radeon 8060S integrated graphics featuring 40 RDNA 3.5 compute units and an AMD XDNA 2 NPU with up to 50 TOPS.
- Linux AI Developer Platform: Purpose-built for Linux-based AI development with full AMD ROCm software support and preloaded tools, models, and workflows optimized for local AI development.
- Compact, Connected Design: Includes a 2TB M.2 SSD, 10GbE LAN, Wi-Fi 7, Bluetooth 5.4, USB-C connectivity, and HDMI 2.1b.
The rate also depends on the performance level being tracked. Averaged across the five benchmarks, Epoch estimates costs at newly achieved state-of-the-art performance fell 66% per quarter (75-fold per year). At the same performance level two years later, the decline was 32% per quarter (4.7-fold per year). In other words, the cost of newly reached frontier performance fell faster than the cost of performance that had become less cutting-edge.
One benchmark example
For GPQA Diamond, Epoch estimates OpenAI o3 could achieve a 75% score for an average of $0.30 per question when it was released on January 31, 2025. The report says GPT-5.6 Luna later reached the same score for $0.0004 per question, a 725-fold reduction in under 18 months. These are the report’s estimates for one benchmark comparison, not general prices for using either model.
How does that compare with Moore’s Law and other technologies?
Epoch compares the rate of price decline in its AI benchmark series with historical price series for other technologies. Its report expresses the AI decline as four times the DNA-sequencing rate, six times the compute rate, 18 times the lithium-ion battery rate, and 54 times the electricity rate. The historical annual decline multipliers below come from different products and time periods, so they are rate comparisons rather than like-for-like price comparisons.
Rank #2
- EVOLUTION AMD RYZEN AI MAX+ 395 MINI PC - GMKtec EVO-X2 is the next evolution in AI mini PC Ryzen Strix Halo series. Thanks to AMD Simultaneous Multithreading (SMT) the core-count is effectively doubled, to 32 threads. Ryzen AI Max+ 395 has 64 MB of L3 cache and can boost up to 5.1 GHz, depending on the workload. The Ryzen AI Max+ 395 is currently rated as the "most powerful x86 APU" on the market for AI computing.
- AI NPU with XDNA 2 ARCHITECTURE - Powered by 16 “Zen 5” CPU cores, 50+ peak AI TOPS XDNA 2 NPU and a truly massive integrated GPU driven by 40 AMD RDNA 3.5 CUs, the Ryzen AI MAX+ 395 is a transformative upgrade and delivers a significant performance boost over the competition. The Ryzen AI Max+ 395 excels in consumer AI workloads like the llama.cpp-powered application: LM Studio. Shaping up to be the must-have app for client LLM workloads, LM Studio allows users to locally run the latest language model without any technical knowledge required and unleash their creativity and productivity.
- AMD RADEON 8090S iGPU GAMING PC - The AMD Radeon RX 8060S offers all 40 CUs with up to 2.9 GHz graphics clock and uses the new RDNA 3.5 architecture. The powerful iGPU is positioned between an RTX 4060 and 4070 laptop GPU and therefore enables gaming in FHD at maximum details in most demanding games. The 8060S can also utilize the full 64GB pool, which is perfect for running LLMs such as Deepseek 32B, which runs comfortably on this machine.
- EIGHT CHANNEL LPDDR5X - LPDDR5X is a new ground breaking memory small form factor installed on-board. With blazing speeds up to to 8000MT/s, it runs 1.5x faster than the DDR5 SODIMMs; 90% better performance over DDR5 SODIMMs in video conferencing and photo editing; 30% better performance in productivity apps; 4% better performance in digital content workloads.
- QUAD SCREEN 8K DISPLAY SUPPORT - EVO-X2 AI Mini PC support 4-screen 4K/8K output via HDMI 2.1 (8K@60Hz), DisplayPort 1.4 (4K@60Hz), and dual USB 4 40Gbps Transfer speed (supporting PD3.0/DP1.4/DATA). Ideal for gaming, video editing, and multitasking, it provides expansive and crisp multi-display support.
| Technology and series | Historical annual decline multiplier reported by Epoch AI | Period examined |
|---|---|---|
| AI benchmark performance costs | About 13× per year | Since 2023, across five benchmarks |
| DNA sequencing | 1.84× per year | 2001–2025 |
| Compute | 1.51× per year | 1940–2001 |
| Lithium-ion batteries | 1.16× per year | 1991–2024 |
| US residential electricity | 1.05× per year | 1892–1973 |
Moore’s Law is often used as shorthand for a long-running trend in computing, but Epoch’s direct comparator is a historical compute price series—not a claim that one law governs the cost of AI today. AI benchmark costs and the historical series involve different units, products and time windows. Epoch itself cautions that the comparison is “apples and oranges”: AI performance is not a single, uniform good.
Does this mean AI is 13 times cheaper for users?
No. The estimate concerns the cost frontier for reaching benchmark scores, not what every customer pays. It assumes someone can find and switch to the cheapest suitable model for a given target. Many users stay with a provider or product for reasons besides price, and a benchmark score may not establish that a model delivers the quality, reliability or features a particular task requires.
Two other analyses add context but measure different things:
Rank #3
- Intel Core Ultra 9 285 Processor: Newly developed cores deliver ultra-smooth and responsive gameplay. AI accelerators prepare users for the next era of gaming on an AI PC.
- Simplistic Design: Enjoy the latest generation of Windows 11 Home for your everyday needs. *MSI recommends Windows 11 Pro for business use.
- NVIDIA GeForce RTX 5070 Ti GPU
- Cool While Gaming: In conjunction with an RGB CPU Air Cooler, the Aegis RS features four system cooling fans; three in the front and one in the rear to pull in cool air and push heat out of the PC.
- Turn on the Bright Lights: With the built-in RGB lighting, take your gaming experience to the next level by pressing the MSI LED button to cycle through lighting options. Customize lighting even further with MSI Center software.
- A 2026 Journal of Economic Perspectives article by Mert Demirer, Andrey Fradkin and Nadav Tadelis describes an LLM inference market with rapid turnover among leading models and creators, and no single model dominating every use case. Its market analysis reports a 1,000-fold decline in the price of intelligence and prices about 90% lower for open-source models than comparable closed-source models in its study. These are separate market findings, not measurements of Epoch’s benchmark frontier. See the American Economic Association article.
- Gartner’s March 25, 2026 forecast projects that provider inference costs for a one-trillion-parameter LLM will be over 90% lower in 2030 than in 2025. That is a forecast, with variation depending on semiconductor assumptions—not an observed result or a promise of equivalent price cuts for customers. Gartner warns that savings may not fully pass through to enterprise customers. See Gartner’s forecast and market commentary.
Why can lower AI costs still mean higher bills?
A falling unit cost does not guarantee that total spending will fall. Gartner says agentic models can require five to 30 times more tokens per task than a standard chatbot. If organizations run many more tasks or use more tokens per task, aggregate inference expense can rise even as the cost of each unit falls. Gartner analyst Will Sommer put the distinction this way: “Chief Product Officers (CPOs) should not confuse the deflation of commodity tokens with the democratization of frontier reasoning.”
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallFor a practical model or service comparison, assess quality on the work you need done, total tokens or cost per completed task, latency and access—not a benchmark price alone. Also distinguish the provider’s inference cost from the price charged to you; the cited analyses do not establish a single best provider.
What are the limits of the “in freefall” claim?
Epoch’s series is short, incomplete and noisy, and it does not cover every model-and-benchmark combination. Benchmark success can be an imperfect proxy for useful work, and systems may be optimized for particular benchmarks. The report’s estimates are most useful as evidence of a rapid change in the cost of reaching certain measured performance levels—not as a universal forecast or guarantee of lower bills.
Epoch summarizes its finding: “The cost of achieving a given level of AI performance has fallen about 47% per quarter, faster than for any other transformative technology in history.” That statement belongs to its specified benchmark comparison. It should not be read as saying all AI services became 13 times cheaper, or that users automatically receive the frontier’s full savings.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
