Skip to content
aiqbench

Capability frontier over time

Each Epoch benchmark, plotted on the release date of every model that's been scored on it. As newer models arrive the ceiling of what's been achieved keeps rising — a real trend, distinct from and often mistaken for individual models improving over time.

BenchmarkGPQA DiamondOTIS Mock AIMEFrontierMathSimpleQA VerifiedSWE-bench VerifiedARC-AGI-2Terminal-BenchHumanity's Last ExamSimpleBench

ARC-AGI-2 — models on arrival

One point per model at its release date. The line traces the best-known score at each moment.

0%25%50%75%100%2024202520262027Model release dateScore (%)gpt-4o-2024-11-20 Released: May 2024 Score: 0.0% Date source: curatedgpt-4o-mini-2024-07-18 Released: Jul 2024 Score: 0.0% Date source: curatedo1-mini-2024-09-12 Released: Sep 2024 Score: 0.8% Date source: curatedgemini-1.5-pro-002 Released: Sep 2024 Score: 0.8% Date source: curatedDeepSeek-R1 Released: Jan 2025 Score: 1.3% Date source: curatedo3-mini-2025-01-31 Released: Jan 2025 Score: 3.0% Date source: curatedgemini-2.0-flash-001 Released: Feb 2025 Score: 1.3% Date source: curatedgrok-3-mini Released: Feb 2025 Score: 0.4% Date source: curatedclaude-3-7-sonnet-20250219 Released: Feb 2025 Score: 0.9% Date source: curatedgpt-4.5-preview-2025-02-27 Released: Feb 2025 Score: 0.8% Date source: curatedLlama-4-Maverick-17B-128E-Instruct Released: Apr 2025 Score: 0.0% Date source: curatedLlama-4-Scout-17B-16E-Instruct Released: Apr 2025 Score: 0.0% Date source: curatedgpt-4.1-nano-2025-04-14 Released: Apr 2025 Score: 0.0% Date source: curatedgpt-4.1-mini-2025-04-14 Released: Apr 2025 Score: 0.0% Date source: curatedgpt-4.1-2025-04-14 Released: Apr 2025 Score: 0.4% Date source: curatedo4-mini-2025-04-16 Released: Apr 2025 Score: 6.1% Date source: curatedo3-2025-04-16 Released: Apr 2025 Score: 6.5% Date source: curatedclaude-opus-4-20250514 Released: May 2025 Score: 8.6% Date source: curatedclaude-sonnet-4-20250514 Released: May 2025 Score: 5.9% Date source: curatedDeepSeek-R1-0528 Released: May 2025 Score: 1.1% Date source: curatedmagistral-medium-2506 Released: Jun 2025 Score: 0.0% Date source: name-suffixgemini-2.5-pro Released: Jun 2025 Score: 4.9% Date source: curatedgemini-2.5-pro-preview-06-05 Released: Jun 2025 Score: 0.0% Date source: curatedo3-pro-2025-06-10 Released: Jun 2025 Score: 4.9% Date source: curatedQwen3-235B-A22B-Instruct-2507 Released: Jul 2025 Score: 1.3% Date source: name-suffixgrok-4-0709 Released: Jul 2025 Score: 16.0% Date source: curatedgpt-5-mini-2025-08-07 Released: Aug 2025 Score: 4.4% Date source: curatedgpt-5-2025-08-07 Released: Aug 2025 Score: 9.9% Date source: curatedgpt-5-nano-2025-08-07 Released: Aug 2025 Score: 2.6% Date source: curatedgrok-4-fast Released: Sep 2025 Score: 5.3% Date source: curatedclaude-sonnet-4-5-20250929 Released: Sep 2025 Score: 13.6% Date source: curatedgpt-5-pro-2025-10-06 Released: Oct 2025 Score: 18.3% Date source: curatedclaude-haiku-4-5-20251001 Released: Oct 2025 Score: 4.0% Date source: curatedclaude-opus-4-5-20251101 Released: Nov 2025 Score: 37.6% Date source: name-suffixgpt-5.1-2025-11-13 Released: Nov 2025 Score: 17.6% Date source: name-suffixgpt-5.2-2025-12-11 Released: Dec 2025 Score: 52.9% Date source: name-suffixgpt-5.2-pro-2025-12-11 Released: Dec 2025 Score: 54.2% Date source: name-suffixgpt-5.2-prokimi-k2.5 Released: Jan 2026 Score: 11.8% Date source: huggingface-createdglm-5 Released: Feb 2026 Score: 4.9% Date source: huggingface-createdMiniMax-M2.5 Released: Feb 2026 Score: 4.9% Date source: huggingface-createdgpt-5.4-pro-2026-03-05 Released: Mar 2026 Score: 83.3% Date source: name-suffixgpt-5.4-prokimi-k3 Released: Jun 2026 Score: 60.4% Date source: huggingface-createdglm-5.2 Released: Jun 2026 Score: 22.8% Date source: huggingface-createdInkling Released: Jul 2026 Score: 36.5% Date source: huggingface-createdinkling-small Released: Jul 2026 Score: 40.1% Date source: huggingface-createddeepseek-v4-flash-0731 Released: Jul 2026 Score: 61.4% Date source: huggingface-createdqwen3.8-27b Released: Aug 2026 Score: 42.4% Date source: huggingface-createddeepseek-v4-pro-0813 Released: Aug 2026 Score: 61.3% Date source: huggingface-createdglm-5.3-flash Released: Aug 2026 Score: 65.8% Date source: huggingface-createdclaude-fable-5-1 Released: Sep 2026 Score: 90.0% Date source: lmarena-first-seenclaude-fable-5-1claude-opus-5 Released: Sep 2026 Score: 90.4% Date source: lmarena-first-seenclaude-opus-5claude-fable-5 Released: Sep 2026 Score: 89.2% Date source: lmarena-first-seengemini-3.1-pro-preview Released: Sep 2026 Score: 77.1% Date source: lmarena-first-seengpt-5.4-2026-03-05 Released: Sep 2026 Score: 74.0% Date source: lmarena-first-seengpt-6-astra Released: Sep 2026 Score: 95.0% Date source: lmarena-first-seengpt-6-astraclaude-opus-4-6 Released: Sep 2026 Score: 69.2% Date source: lmarena-first-seengrok-4.6 Released: Sep 2026 Score: 67.1% Date source: lmarena-first-seengpt-5.6-terra Released: Sep 2026 Score: 83.9% Date source: lmarena-first-seengpt-5.5 Released: Sep 2026 Score: 85.0% Date source: lmarena-first-seengpt-5.6-luna Released: Sep 2026 Score: 59.5% Date source: lmarena-first-seenclaude-opus-4-7 Released: Sep 2026 Score: 75.8% Date source: lmarena-first-seengpt-5.6-sol Released: Sep 2026 Score: 92.5% Date source: lmarena-first-seengrok-4.5 Released: Sep 2026 Score: 52.6% Date source: lmarena-first-seenclaude-sonnet-4-6 Released: Sep 2026 Score: 60.4% Date source: lmarena-first-seengemini-3.6-flash Released: Sep 2026 Score: 60.4% Date source: lmarena-first-seengemini-3.5-flash Released: Sep 2026 Score: 72.1% Date source: lmarena-first-seenclaude-opus-4-8 Released: Sep 2026 Score: 72.1% Date source: lmarena-first-seengemini-3.8-flash Released: Oct 2026 Score: 89.2% Date source: lmarena-first-seengpt-5.4-mini-2026-03-17 Released: Oct 2026 Score: 18.9% Date source: lmarena-first-seengpt-6-luna Released: Oct 2026 Score: 59.3% Date source: lmarena-first-seengpt-5.4-nano-2026-03-17 Released: Oct 2026 Score: 5.7% Date source: lmarena-first-seengemini-3.5-flash-lite Released: Oct 2026 Score: 10.3% Date source: lmarena-first-seengpt-6.1-sol Released: Oct 2026 Score: 94.2% Date source: lmarena-first-seengpt-6-sol Released: Oct 2026 Score: 89.6% Date source: lmarena-first-seenclaude-opus-5-5 Released: Oct 2026 Score: 93.3% Date source: lmarena-first-seengemini-3.7-flash Released: Oct 2026 Score: 84.6% Date source: lmarena-first-seengrok-4.7 Released: Oct 2026 Score: 61.4% Date source: lmarena-first-seendeepseek-v4.1-flash Released: Oct 2026 Score: 72.9% Date source: huggingface-created
Model at its release dateRunning best score (frontier)Score confidence interval

Read this as a frontier chart, not a per-model trajectory. A fixed model's score on a fixed benchmark does not change over time — Claude 3.5 Sonnet's GPQA number today is the same one Anthropic reported at launch. The line rises because better models keep arriving, not because any specific model got better. Seeing one model's score change over time (as its evaluation is re-run with different scaffolds, for example) needs a different kind of data: daily snapshots of each board, which we will chart once several weeks of them have accumulated.

21 models have no known release date and are not plotted.

Benchmarks · frontier

Frontier milestones

Every time a ceiling moved: who, which benchmark, by how much. Tap a sparkline to pick a benchmark — the chart above follows it; tap an organisation to filter by lab. Each event is the diff of the running best the chart draws.

16 of 114 frontier events ·

  1. Sep 202690.0%

    claude-fable-5-1 Anthropic sets ARC-AGI-2

    +6.7 pp
  2. Sep 202690.4%

    claude-opus-5 Anthropic sets ARC-AGI-2

    +0.4 pp
  3. Sep 202695.0%

    gpt-6-astra OpenAI sets ARC-AGI-2

    +4.6 pp
  4. Mar 202683.3%

    gpt-5.4-pro-2026-03-05 OpenAI sets ARC-AGI-2

    +29.2 pp
  5. Dec 202552.9%

    gpt-5.2-2025-12-11 OpenAI sets ARC-AGI-2

    +15.3 pp
  6. Dec 202554.2%

    gpt-5.2-pro-2025-12-11 OpenAI sets ARC-AGI-2

    +1.3 pp
  7. Nov 202537.6%

    claude-opus-4-5-20251101 Anthropic sets ARC-AGI-2

    +19.3 pp
  8. Oct 202518.3%

    gpt-5-pro-2025-10-06 OpenAI sets ARC-AGI-2

    +2.4 pp
  9. Jul 202516.0%

    grok-4-0709 xAI sets ARC-AGI-2

    +7.4 pp
  10. May 20258.6%

    claude-opus-4-20250514 Anthropic sets ARC-AGI-2

    +2.1 pp
  11. Apr 20256.1%

    o4-mini-2025-04-16 OpenAI sets ARC-AGI-2

    +3.1 pp
  12. Apr 20256.5%

    o3-2025-04-16 OpenAI sets ARC-AGI-2

    +0.4 pp
  13. Jan 20253.0%

    o3-mini-2025-01-31 OpenAI sets ARC-AGI-2

    +1.7 pp
  14. Jan 20251.3%

    DeepSeek-R1 DeepSeek sets ARC-AGI-2

    +0.5 pp
  15. Sep 20240.8%

    o1-mini-2024-09-12 OpenAI sets ARC-AGI-2

    +0.8 pp
  16. May 20240.0%

    gpt-4o-2024-11-20 OpenAI sets ARC-AGI-2

    first score