Artificial Analysis Intelligence Index v4.1
Version 4.1 of Artificial Analysis's index, which moved to GDPval-AA v2, Terminal-Bench 2.1 and τ³-Banking and put the most weight on agents.
In use from June to September 2026, with 4.1.1 (August 2026) changing the τ³-Banking dataset and the graders for HLE, AA-LCR and AA-Omniscience. Components: GDPval-AA v2, τ³-Banking, Terminal-Bench 2.1, SciCode, AA-LCR, AA-Omniscience (accuracy 8%, non-hallucination 4%), Humanity's Last Exam, GPQA Diamond and CritPt. Weights: Agents 34%, Coding 24%, Scientific Reasoning 24%, General 18%. IFBench left the index.
- Scored
- Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
- Answers
- Agent task Many steps in a terminal, browser, desktop, app, game or robot.
Grok 4.6 leads with 61.0 among models released by Aug 2026
Best score per model as labs report them in their own cards. Labs pick their own settings, so these compare less cleanly than independent runs.
The newest model with a score here came out in Aug 2026, and nobody has reported a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6 Astra
- Claude Opus 5.5
- GPT-6.1 Terra
- GPT-6.1 Sol
- and 13 more
- 1 Grok 4.6 61.0
- 2 GPT-5.6 Sol 58.9
- 3 Kimi K3 57.1
- 4
Gemini 3.7 Flash 56.0 - 5 GPT-5.6 Terra 55.0
- 6 GPT-5.6 Luna 51.2
- 7
Inkling-Small 40.0
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | Grok 4.6 | 61.0 | Aug 2026 | Model card | |
| 2 | GPT-5.6 Sol | 58.9 | Jul 2026 | Model card | |
| 3 | Kimi K3 | 57.1 | Jul 2026 | Model card | |
| 4 | Gemini 3.7 Flash | 56.0 | Aug 2026 | Model card | |
| 5 | GPT-5.6 Terra | 55.0 | Jul 2026 | Model card | |
| 6 | GPT-5.6 Luna | 51.2 | Jul 2026 | Model card | |
| 7 | Inkling-Small | 40.0 | Jul 2026 | Model card |
2 models have held number one since its release in Jun 2026
Lab claims count from the day the card came out, independent runs from the model’s release.
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
4 labs put it in a launch post in the last six months
Launch posts from every lab that has listed it, one tick per post. Tall colored ticks include it, and short gray ones leave it out.
| Date | Launch post | Score | As the post writes it |
|---|---|---|---|
| 13 Aug 2026 | Gemini 3.7 Flash | 56 | Artificial Analysis Intelligence Index |
| 12 Aug 2026 | Grok 4.6 | 61 | AA Intelligence Index |
| 30 Jul 2026 | Inkling-Small | 40 | AA Index v4.1 |
| 9 Jul 2026 | GPT-5.6 Luna | 51.2 | Artificial Analysis Intelligence Index v4.1 |
| 9 Jul 2026 | GPT-5.6 Terra | 55 | Artificial Analysis Intelligence Index v4.1 |
| 9 Jul 2026 | GPT-5.6 Sol | 58.9 | Artificial Analysis Intelligence Index v4.1 (Sol, max) |
3 system cards report it, from 3 labs
Besides launch posts, we read the system cards, model cards and technical reports labs publish with a release. They report far more, including safety and frontier-risk tests.
| Date | Document | Score | As the document writes it | Section |
|---|---|---|---|---|
| 13 Aug 2026 | Gemini 3.7 Flash model card | 56 | Artificial Analysis Intelligence Index (results table, p.5) | Capabilities |
| 30 Jul 2026 | Inkling-Small model card | 40 | AA Index v4.1 (5. Evaluations table) | Capabilities |
| 27 Jul 2026 | Kimi K3 technical report | 57.1 | Artificial Analysis Intelligence Index v4.1, rank 4 of 580 (Table 5) | Other |
Versions
More in Other
- Personal Superintelligence Index (Meta) In use Meta's composite of seven personal and knowledge-work evals: Agents' Last Exam, AutomationBench-AA, FocusBench, JobBench, GDP.pdf, VBT and DRACO.
- Anthropic ECI (AECI) Only in system cards Index that combines a basket of capability benchmarks into one capability score to track trends across Anthropic models. Claude Opus 5.5 169.4
- Qwen in-house English automatic evaluation Only in system cards Models answer Qwen in-house English datasets on knowledge, comprehension, coding, math and related skills. Skills and weights change between reports.
Qwen2-72B 72.9% - Recognition of model-written user turns (Anthropic) Only in system cards Models write synthetic user turns for real Claude Code transcripts, then judge which version is real. Score is accuracy above chance. Claude Mythos Preview 0.35
- LiveBench Dropped Models answer monthly refreshed questions across math, coding, reasoning, language, instruction following, and data analysis. Scored automatically against objective ground truth. Claude Fable 5.1 85.7%
- PhiBench Dropped Models answer original prompts that Microsoft's Phi team wrote to test a wide range of skills and reasoning.
Phi-4-Reasoning-plus 74.2%