How long a task can an AI agent finish on its own?
AI task horizon at 50% reliability (METR) · minutes · measured 2019–2026 · fitted live from 7 observations by the Many Minded cost-curve engine.
The last measured value is 17.4 hr (2026, METR Horizon v1.1 (official measurements)). Over the fitted window the series rises 314.1% a year (80% interval +172.9% to +528.5%) — a doubling every 0.5 years (80%: 0.4–0.7). Carried forward on that fit, 2028: 299 hr (80%: 50.9 hr–1.75e+3 hr).
the curve
what this series measures
length of human task (in minutes) that frontier AI agents complete at 50% reliability — the standard capability driver series; yearly SOTA.
the gate — what crossing this threshold changes
Central fit: ~2027. 80% window: 2027–2028 · n=7. That is 1.2 doublings away at today's value.
what the fit says
| quantity | value | how it is computed |
|---|---|---|
| fitted rate | +314.1%/yr (80%: +172.9% to +528.5%) | Farmer–Lafond drift over the trailing window · n=7 points, 7 years |
| doubling time | 0.5 years (80%: 0.4–0.7) | implied by the fitted drift |
Not shown for this curve, because the machinery returns nothing: Wright's law (no cumulative-deployment series for this technology); the accelerating/decelerating check (needs ≥8 observations, this has 7); a regime break (needs ≥12 observations).
the projection, with its interval
Log cost as a random walk with drift: the forecast variance grows with horizon (τ + τ²/m), which is why these bands widen instead of staying parallel. The middle column is the least useful number on this page; the interval is the claim.
| year | central fit | 80% interval | |
|---|---|---|---|
| 2027 | 72.1 hr | 22.2 hr to 235 hr | near horizon |
| 2028 | 299 hr | 50.9 hr to 1.75e+3 hr | near horizon |
| 2029 | above 400 hr | — | past the point where a number would be theater |
The table stops at 400 hr — a decade above the gate. Past that, quoting a number would be theater rather than forecast.
questions this page answers
How long a task can an AI agent finish on its own?
17.4 hr as of 2026, the latest measured value in the series (METR Horizon v1.1 (official measurements)). The fitted trend has it rising 314.1% a year, with an 80% interval of +172.9% to +528.5%.
How fast is the AI task horizon rising?
+314.1% a year over the fitted window, an 80% interval of +172.9% to +528.5% — a doubling every 0.5 years (80%: 0.4 to 0.7 years). Fitted from 7 observations spanning 7 years.
What will the AI task horizon be in 2028?
The central fit says 299 hr, inside an 80% interval of 50.9 hr to 1.75e+3 hr. The interval is the forecast; the middle number is only its midpoint. Bands widen with horizon because shocks accumulate — a constant-width band would be overconfident.
When will the AI task horizon reach ≥ 40-hour task horizon?
The central fit says ~2027, with an 80% window of 2027–2028 from 7 observations. Crossing it is what the engine calls "AI agents hold week-long work", firing: R&D everywhere — the master trigger. Treat the window, not the year, as the claim.
Where does this data come from?
METR Horizon v1.1 (official measurements). 7 observations spanning 2019–2026. Curated benchmark history, extended by a weekly authoritative fetch and by news figures fact-checked against their source before they may touch a fit. Both the observation ledger and the fitting code are public, and the engine publishes its own calibration score and its misses.
the raw numbers
Every observation behind the fit, unrounded by us and unsmoothed. This table is here on purpose: graphs make people underestimate exponential change, and the raw series beside the curve is the one correction shown to work.
| year | minutes | change |
|---|---|---|
| 2019 | 0.0500 min | — |
| 2020 | 0.140 min | +180.0% |
| 2022 | 0.600 min | +107.0%/yr |
| 2023 | 3.99 min | +565.0% |
| 2024 | 38.8 min | +873.2% |
| 2025 | 5.87 hr | +807.2% |
| 2026 | 17.4 hr | +196.6% |
where this comes from
- Source: METR Horizon v1.1 (official measurements).
- Basis: measured. Curated benchmark history, maintained by hand against the source, extended by figures our daily editor extracts from the news — each of which is fact-checked against its source article before it may touch a fit.
- Observations: n=7, spanning 2019–2026. Ledger last updated 2026-11-02.
- Fit: the Farmer–Lafond stochastic model — log cost as a random walk with drift over a 10-year trailing window, Student-t tails, and (for series under 15 points) a volatility floor at their drift-volatility prior. The code is readable source; the observation ledger is public JSON.
- Reuse: this compiled series and the numbers on this page are published under CC BY 4.0 — take them, cite the page. The underlying source keeps its own terms.
- Accuracy: the engine grades its own intervals in public — every feasible hindcast re-fitted on data as it stood, scored for whether the actual landed inside the 80% band. The calibration score is on the engine page, misses included.
the rest of the graph
Other KNOWLEDGE curves:
- Hard drive cost per terabyte $/TB
- Transistors per microprocessor — Moore's law count
- Fastest supercomputer, computational capacity GFLOP/s
- DRAM memory cost per terabyte $/TB
- Solid-state storage cost per terabyte $/TB
- Frontier AI training compute, FLOP FLOP
- Frontier-grade AI inference price $/Mtok
Recent briefs from KNOWLEDGE, the daily record of what moved:
- A detector that reads a paper's argument structure to spot AI authorship 2026-11-08University of Maryland researchers built IdeaLens, which flags AI-written text by the shape of its ideas, not its wording. It is still in development.
- Ripple-65B Preview: Open-Source Coding Model Shows Promise But Needs Refinement 2026-11-07Ricursive Intelligence released a 65-billion-parameter coding model with strong benchmark results, but it remains a preview model not yet ready for production use.
- AI research agents still fail to meet expert standards in autonomous scientific work 2026-11-07Nine AI agents tested on open research questions from US national labs failed to meet experts' standards for credible scientific output
- Schools get Gemini switched on first, and a real off-switch only after pushback 2026-11-07Google enabled Gemini for 150 million Classroom students, then split its controls after criticism. EFF argues the default still favors Google, not schools.
- Chinese open-weight models become a default pick for enterprises going open 2026-11-06Low-cost open-weight models from Chinese labs, Moonshot AI's Kimi among them, are reportedly leading enterprise open-source AI spending.
- X-ray evidence points to altermagnetism in manganese telluride 2026-11-06A SLAC-led team reports direct evidence of handedness in manganese telluride's magnetic structure, supporting a third class of magnetism.
Every tracked curve: utility solar, installed · battery pack price · battery cell price · solar module price · human genome sequencing · onshore wind electricity · solar electricity (LCOE) · space launch, to LEO · wheat yield · offshore wind electricity · modular housing, turnkey · AI video, finished minute · humanoid robot, capable unit · cheapest humanoid, list price · people still offline · people outside a mobile-broadband signal · farm labor share · healthcare spend per person
← the whole engine: build clock, cascade, calibration · the archive · how we know