The Moving Pareto Frontier of Cost per Task for LLMs
When you pick a model for an application you are trading off how capable it is against how much each call costs. When using agents to write code, I have focused almost entirely on capability, with the general sense that the improved quality of the output of the work is worth the money, even when far cheaper models exist that are reasonably capable. Recent projects have drawn me to evaluate the full distribution of this trade-off, and I have been shocked by the progress across the cost spectrum, particularly how intelligent cheap models have become.
Artificial Analysis has been benchmarking intelligence and price across hundreds of models for a couple of years, and enough of that data is accessible to reconstruct the tradeoff. In particular, this plot shows the intelligence index vs. the cost per task, providing a realistic cost estimate for different levels of model capability. The top line is what they define as the “Pareto line,” the most capable models at a given price point. This line describes the true frontier of LLMs. I pulled data from artificialanalysis.ai and looked at how the Pareto frontier has moved as new models have been released. I think it is worthwhile to take a beat to review this progress and make some predictions for the next few months.
The short version: for a fixed capability level, cost is collapsing. A task at Intelligence Index 50, a level first reached in this data by GPT-5.4 in March at $1.10 per task, could be had for $0.03 by July.
The Artificial Analysis Intelligence Index
The capability axis throughout this post is the Artificial Analysis Intelligence Index, so it is worth being clear about what that number is. The current version, v4.1.1, is a weighted average over nine evaluations grouped into four categories: agentic tasks at 34%, coding at 24%, scientific reasoning at 24%, and general capability at 18%. The weighting reflects where the field’s attention is: a third of the score comes from a model’s ability to complete multi-step agentic work, not from answering exam questions. The component evaluations, their weights, and the scoring details are documented in Artificial Analysis’s intelligence benchmarking methodology.
What you end up with is a single number that represents model capability, sort of like an IQ for LLMs. It isn’t perfect, and two models with the same score may have different strengths, but I have found that this score does a reasonably good job of indicating a model’s capability.
As a reference point, Anthropic’s “Claude 4.5 Sonnet (Reasoning)” was for me and many others the first time a model felt capable enough to use in an agentic harness for writing code. At the time I was using Cline, and this model provided substantial productivity gains over auto-complete and copy/paste workflows. That model had an intelligence score of 37.4 (based on today’s intelligence scoring system). The top current model is Claude Opus 5 max effort, at 63.1.
Measuring Cost per Task
Cost per token is easily available, but different models can use a very different number of tokens, so a better indication of the cost of a model needs to take this into account. Cost per task is Artificial Analysis’s own measured number: the average cost in USD to run one task from their Intelligence Index evaluation suite, including the input, reasoning, and answer tokens actually billed during the run. The website displays it but the free API tier does not include it, so I scraped it from the data embedded in each model’s page on the site, covering both the models they currently benchmark and retired models whose pages still carry the measurement (older Claude Opus and Sonnet versions, the GPT-5.x line, and others). That yields measured cost for 137 models reaching back to DeepSeek V3 in December 2024, each paired with a release date and an Intelligence Index score on the current scale.
How the Frontier Has Moved
The chart below plots intelligence against measured cost per task and traces the Pareto frontier, the cheapest way to reach each intelligence level, as it stands today and as it stood at two month intervals over the past year, using each model’s release date to reconstruct what was available. Hover any point for the model behind it.
Each successive frontier sits above and to the left of the last: more intelligence at the same cost, or the same intelligence for less. The right edge tells the capability story. The ceiling of the frontier rose from index 35.3 in August 2025 (GPT-5 at $0.26 per task) to 37.4 that October (Claude 4.5 Sonnet), 48.4 in February (Claude Sonnet 4.6), 55.0 in April (Claude Opus 4.7 at $2.23), 62.1 in June (Claude Fable 5 at $3.14), and 63.1 today (Claude Opus 5 at $2.34): twenty eight Intelligence Index points in a year. The left half shows the rising intelligence of cheap models. As of August 19, 2026, the GPT-5.6 Luna effort ladder now owns almost everything below index 52, with the level that was the August 2025 ceiling available for $0.0088 per task.
Reasoning effort controls mean a single model is not a point on this chart but a segment: the GPT-5.6 Luna ladder runs from $0.0088 at low effort to $0.047 at max and covers the whole lower half of the frontier, while Claude Opus 5 spans $0.43 at low effort to $2.34 at max and buys about ten Intelligence Index points along the way. Effort is now a key dial in the cost/intelligence trade-off, and this dial requires that any analysis of capability must also consider cost.
The records view tracks the cheapest measured cost per task achieved by any released model at or above a given Intelligence Index tier. The series reaches back to mid 2025 for the lower tiers, and higher tiers appear when they become available.
| Tier | First measured crossing | Cost collapse | Halving time |
|---|---|---|---|
| Index ≥ 30 | Aug 2025 (GPT-5 high) | 29x | ~69 days |
| Index ≥ 40 | Feb 2026 (Claude Sonnet 4.6) | 56x | ~24 days |
| Index ≥ 50 | Mar 2026 (GPT-5.4 xhigh) | 35x | ~25 days |
| Index ≥ 60 | Jun 2026 (Claude Fable 5 max) | 3.8x | ~34 days |
A capability level is first reached by a large frontier model at a premium price. After some time, cheaper models arrive at the same level, and the record steps down by an order of magnitude or more. The ≥ 40 tier opens with Claude Sonnet 4.6 in February at $1.22 per task, undercut within two days by Gemini 3.1 Pro Preview at $0.33; MiMo-V2.5-Pro, an open weights model, cut the record to $0.034 in April, and GPT-5.6 Luna on high effort holds it at $0.022 today. The ≥ 50 tier follows the same arc a month behind: GPT-5.4 crossed it in March at $1.10, GPT-5.5 and then GLM-5.2 and Grok 4.5 walked the record down through the spring, and GPT-5.6 Luna’s xhigh setting reached it at $0.032 in July, a 35 fold drop in four months.
The mechanism is visible in the model names that set each record. The first crossing is a maximum effort frontier model priced at launch premium, most often from Anthropic or OpenAI. Following this, small distilled models from the big labs (the GPT-5.6 Luna line holds three of the four current records), and open weights releases (MiMo, DeepSeek V4, GLM, Hy3) drive rates down dramatically. Across the tiers with enough history to measure, the records halve roughly every three to ten weeks.
The top tier is where the premium survives. Only six models score 60 or above, and the cheapest of them, Grok 4.6, still costs $0.84 per task. But that record has fallen 3.8x since June, and if the pattern from lower tiers holds, a distilled model at this level should collapse the price within a couple of quarters.
To me, the most impressive result is the low price of OpenAI’s “GPT-5.6 Luna” given its intelligence. Now, the intelligence of Anthropic’s “Claude 4.5 Sonnet (Reasoning)” that set off the coding harness revolution 10 months ago is available using “GPT-5.6 Luna (medium)” for 1/40th the cost!
Caveats
The most important limitation is that costs are the latest measured values indexed by release date, not historical measurements taken at release. Prices get cut over a model’s life, so early points reflect any cuts since launch, which biases the analysis toward understating the collapse. Coverage is the second issue. Retired models are included only when their pages still carry the measurement, which recovered 44 of 228 retired models with prices and scores; the rest, o3, GPT-5.3 Codex, and everything from the GPT-4 era among them, are invisible, so the oldest frontiers rest on fewer models than actually existed and the true opening price of the lower tiers was likely set by models this analysis cannot see.
The retired models’ Intelligence Index scores are on the current scale, but the Index itself is one aggregate of many evaluations. And cost per task on an evaluation suite is a reasoning heavy workload with long prompts; a chat workload with short prompts and short answers would scale differently across models, particularly between reasoning and non-reasoning variants.
Predictions
Extrapolating measured rates is risky, since each collapse is a competition event rather than a law, but the arcs have been regular enough to be worth putting numbers on. At the ≥ 60 tier’s current halving time of about 34 days, Grok 4.6’s $0.84 record falls below ten cents around the start of December. Index 55, which Grok 4.5 holds at $0.36 today, should cost under a dime by mid October. The ceiling is harder to call: it climbed twenty eight points over the year but only one point since June, which reads as saturation of the current index rather than a slowdown in the models, so I expect the next milestone there to be an index revision rather than a big number. And if the pattern of the last four tiers holds, whatever the revised index calls the frontier will debut at a few dollars per task and be commoditized within a quarter.
What I Take from This
The motion of the frontier is more predictable than any individual release. Every capability tier so far has followed the same arc: premium debut, rapid commoditization, a settled record held by a distilled or open weights model at a few percent of the debut price. If a capability exists at any price today, the sensible planning assumption is that it will exist at commodity price within months. For system design, that argues for architectures where the model is a swappable component and the routing between capability tiers is explicit, because the tier boundaries themselves have not settled and show no sign of settling soon.
This rapid price reduction opens new applications. Cost-sensitive developers, particularly in economically disadvantaged regions, can now run capable coding agents at a tiny fraction of last year’s cost. Workloads that were priced out entirely a year ago are becoming routine: at GPT-5.6 Luna prices, a thousand tasks cost about ten dollars, and scanning thousands of documents, like scientific literature or legal records, is becoming more affordable every day.
Model metadata pulled from the Artificial Analysis free API, and measured cost per task scraped from the model pages on artificialanalysis.ai, on August 19, 2026. Corrections welcome.