selfdriven AI / Situational Awareness Framework

Research summary · March 2026

AI capability progression 2024–2026

Frontier AI capability has advanced faster in 2024–2026 than in any prior two-year period. Three benchmarks illustrate the pace — and the acceleration is real.

Benchmark analysis · Frontier models · Sources: Epoch AI · Stanford HAI · METR · ARC Prize

METR · TASK HORIZON Task length doubling every 5–7 months
01 — Overview

Two years of accelerating progress

Frontier AI capability has advanced faster in 2024–2026 than in any prior two-year period. The headline numbers:

80%+

SWE-bench Verified — from 4% in early 2023

94%

GPQA Diamond — from ~36% in early 2024

~$1

Cost per million tokens — from $30+ in 2023 (a 30× drop)

77%+

ARC-AGI-2 — from below 5% at launch in March 2025

Acceleration confirmed. Epoch AI found their Capabilities Index grew almost twice as fast over the last two years compared to the prior two — a ~90% acceleration beginning April 2024, coinciding with the rise of reasoning models and reinforcement learning at frontier labs.

02 — Benchmark trends

Capability across three key metrics

SWE-bench (coding), GPQA Diamond (scientific reasoning) and AIME math competition scores — each showing consistent upward progression from March 2024 to March 2026.

Benchmark score progression

Best-in-class scores per benchmark, Mar 2024 – Mar 2026

  • SWE-bench (coding)
  • GPQA Diamond (science)
  • AIME math
Data sourced from model cards, Stanford HAI AI Index 2025, and Epoch AI Benchmarking Hub. Scores reflect best published results per period across all frontier models.
03 — METR time horizon

Task length doubling every 5–7 months

METR measures the length of software engineering tasks an AI can complete with a 50% success rate. The log scale shows the exponential trend clearly — the linear view reveals the dramatic acceleration.

AI task horizon (log scale)

Exponential growth appears as a straight line on a log axis

  • Measured task length
  • 7-month doubling trend
The ringed point marks November 2025: Claude Opus 4.5 completing ~5-hour human tasks. Doubling pace accelerated from every 7 months (2024) to every ~5 months by mid-2025.

AI task horizon (linear scale)

The same data without log compression — the hockey stick shape

  • Measured task length
  • 7-month doubling trend
Linear scale makes the exponential shape visceral — the last two data points account for most of the total growth. This is what 9× improvement in 12 months looks like.
04 — ARC-AGI-2

The inflection benchmark

ARC-AGI-2 launched in March 2025 with all frontier models below 5%. Within a year, scores surged past the human baseline of 60%. This is the sharpest capability inflection on record.

ARC-AGI-2 score progression

Best AI score vs. human baseline (60%) — Mar 2025 to Mar 2026

  • Best AI score
  • Human baseline
ARC-AGI-2 tests abstract reasoning that can't be brute-forced — easy for humans, hard for AI. Progress driven by refinement loops and test-time compute. Sources: ARC Prize Foundation, arcprize.org.

What changed: Progress wasn't just larger models — it was application-layer engineering. Refinement loops and structured test-time reasoning pushed scores from near-zero to beyond the human baseline in under 12 months. ARC-AGI-3 (interactive reasoning) launched March 2026 to reset the challenge.

05 — Key findings

What the data shows

Six observations from two years of frontier AI development.

06 — Model timeline

Key releases & inflection points

The models that moved the benchmarks, in order.

  1. Mar 2024

    Claude 3 Opus / GPT-4 Turbo

    Last generation before the reasoning model era. GPQA ~40%, SWE-bench ~18%.

  2. Jun 2024

    GPT-4o

    Multimodal as default. Real-time voice. Cost drop cycle begins.

  3. Sep 2024

    Claude 3.5 Sonnet

    New SWE-bench record. Computer use introduced. SWE-bench ~38%.

  4. Nov 2024

    OpenAI o1 + DeepSeek V3

    Test-time compute arrives. o1 scores 74% on IMO qualifier vs GPT-4o's 9%. DeepSeek signals China's capability.

  5. Mar 2025

    ARC-AGI-2 launched

    All frontier models score below 5%. Human baseline: 60%. The hockey stick benchmark begins.

  6. May 2025

    Claude 4 / GPT-4.5

    Reasoning as standard, not premium. GPQA crosses 80%+. SWE-bench ~64%.

  7. Aug–Nov 2025

    Gemini 3 / Claude 4.5 / Grok 4

    Claude Opus 4.5 completes 5-hour software tasks. ARC-AGI-2 breaks 45%. Gemini 3 hits 100% AIME 2025.

  8. Mar 2026

    Claude 4.6 / Gemini 3.1 Pro / GPT-5.4

    ARC-AGI-2 at 77–84%+. SWE-bench 80%+. 1M context windows. ARC-AGI-3 (interactive) released.

selfdriven.ai · Compiled March 2026

Sources: Epoch AI · Stanford HAI AI Index 2025 · METR · ARC Prize Foundation · LXT · llm-stats.com

Data reflects best published scores across frontier models per period. Some scores estimated from trend data.