AI trends
What actually changed
Two things move at once: the ceiling keeps rising, and the floor price of any given capability keeps collapsing. The second one changes more decisions than the first.
Frontier record holder
63.1
Claude Opus 5 · Jul 2026
Index points gained
+56.3
since May 2023
Price of Index 40+
44× cheaper
since it first appeared
Frontier records set
18
distinct record holders
The ceiling
Frontier intelligence over time
Best Intelligence Index score available on any given date. Flat stretches are the interesting part — that is how long a record held.
Frontier Intelligence Index
Single series · a step means a new record holder
Hover the plot for the model holding the record at that date.
View as table
| Date | Model | Index |
|---|---|---|
| May 2023 | GPT-4 (OpenAI) | 6.8 |
| Apr 2024 | GPT-4 Turbo (OpenAI) | 7.7 |
| May 2024 | GPT-4o (OpenAI) | 11.1 |
| Dec 2024 | o1 (OpenAI) | 23.9 |
| Apr 2025 | o4 Mini (OpenAI) | 26.1 |
| Apr 2025 | o3 (OpenAI) | 31.1 |
| Jun 2025 | o3 Pro (OpenAI) | 33.3 |
| Aug 2025 | GPT-5 (OpenAI) | 35.3 |
| Nov 2025 | GPT-5.1-Codex (OpenAI) | 35.6 |
| Nov 2025 | GPT-5.1 (OpenAI) | 37.5 |
| Dec 2025 | GPT-5.2 (OpenAI) | 43.3 |
| Feb 2026 | Gemini 3.1 Pro Preview (Google) | 47.7 |
| Mar 2026 | GPT-5.4 (OpenAI) | 53.1 |
| Apr 2026 | Claude Opus 4.7 (Anthropic) | 55.0 |
| Apr 2026 | GPT-5.5 (OpenAI) | 56.3 |
| May 2026 | Claude Opus 4.8 (Anthropic) | 57.3 |
| Jun 2026 | Claude Fable 5 (Anthropic) | 62.1 |
| Jul 2026 | Claude Opus 5 (Anthropic) | 63.1 |
The floor
What a capability costs over time
Cheapest blended price at which each intelligence tier was available. Each line steps down when somebody undercuts the previous best.
Cheapest price per intelligence tier
Blended USD per 1M tokens on a log scale · lower is better
- Index 30+
- Index 40+
- Index 55+
A tier only appears once some model reached it. Log scale, because the drops are order-of-magnitude. Where one model is the cheapest option for two tiers at once, those lines coincide — hover to read every tier at a date.
View as table
| Tier | Date | Model | Blended / 1M |
|---|---|---|---|
| Index 30+ | Apr 2025 | o3 | $3.5 |
| Index 30+ | Aug 2025 | GPT-5 | $3.44 |
| Index 30+ | Nov 2025 | Kimi K2 Thinking | $1.07 |
| Index 30+ | Nov 2025 | GPT-5.1-Codex-Mini | $0.688 |
| Index 30+ | Dec 2025 | M2.1 | $0.525 |
| Index 30+ | Feb 2026 | M2.5 | $0.39 |
| Index 30+ | Apr 2026 | MiMo-V2.5 | $0.175 |
| Index 30+ | Apr 2026 | Hy3 preview | $0.1 |
| Index 30+ | Jul 2026 | Ling-3.0-flash | $0.032 |
| Index 40+ | Dec 2025 | GPT-5.2 | $4.81 |
| Index 40+ | Feb 2026 | GLM 5 | $1.35 |
| Index 40+ | Apr 2026 | Qwen3.6 Plus | $0.731 |
| Index 40+ | Apr 2026 | MiMo-V2.5-Pro | $0.544 |
| Index 40+ | Apr 2026 | V4 Flash 0423 | $0.11 |
| Index 55+ | Apr 2026 | Claude Opus 4.7 | $10 |
| Index 55+ | Jun 2026 | Claude Sonnet 5 | $4 |
| Index 55+ | Jul 2026 | Grok 4.5 | $3 |
| Index 55+ | Jul 2026 | GPT-5.6 Terra | $2.25 |
| Index 55+ | Aug 2026 | Muse Spark 1.2 | $2 |
The race
Six creators, best-so-far
Each line is one creator's own record over time. Convergence at the right edge is the story of the last year.
Best Intelligence Index per creator
Six series · each line steps when that creator beats its own record
- Anthropic
- OpenAI
- Qwen
- DeepSeek
- Meta
Only creators with more than one benchmarked release appear. Lines are direct-labelled at the right edge.
View as table
| Creator | Date | Model | Index |
|---|---|---|---|
| Anthropic | Mar 2024 | Claude 3 Haiku | 3.5 |
| Anthropic | May 2025 | Claude Sonnet 4 | 26.0 |
| Anthropic | Aug 2025 | Claude Opus 4.1 | 28.8 |
| Anthropic | Sep 2025 | Claude Sonnet 4.5 | 29.9 |
| Anthropic | Nov 2025 | Claude Opus 4.5 | 35.6 |
| Anthropic | Feb 2026 | Claude Opus 4.6 | 38.8 |
| Anthropic | Apr 2026 | Claude Opus 4.7 | 55.0 |
| Anthropic | May 2026 | Claude Opus 4.8 | 57.3 |
| Anthropic | Jun 2026 | Claude Fable 5 | 62.1 |
| Anthropic | Jul 2026 | Claude Opus 5 | 63.1 |
| OpenAI | May 2023 | GPT-4 | 6.8 |
| OpenAI | Apr 2024 | GPT-4 Turbo | 7.7 |
| OpenAI | May 2024 | GPT-4o | 11.1 |
| OpenAI | Dec 2024 | o1 | 23.9 |
| OpenAI | Apr 2025 | o4 Mini | 26.1 |
| OpenAI | Apr 2025 | o3 | 31.1 |
| OpenAI | Jun 2025 | o3 Pro | 33.3 |
| OpenAI | Aug 2025 | GPT-5 | 35.3 |
| OpenAI | Nov 2025 | GPT-5.1-Codex | 35.6 |
| OpenAI | Nov 2025 | GPT-5.1 | 37.5 |
| OpenAI | Dec 2025 | GPT-5.2 | 43.3 |
| OpenAI | Feb 2026 | GPT-5.3-Codex | 45.5 |
| OpenAI | Mar 2026 | GPT-5.4 | 53.1 |
| OpenAI | Apr 2026 | GPT-5.5 | 56.3 |
| OpenAI | Jul 2026 | GPT-5.6 Sol | 60.9 |
| Jun 2025 | Gemini 2.5 Pro | 25.9 | |
| Feb 2026 | Gemini 3.1 Pro Preview | 47.7 | |
| May 2026 | Gemini 3.5 Flash | 52.0 | |
| Qwen | Sep 2024 | Qwen2.5 72B Instruct | 9.4 |
| Qwen | Jul 2025 | Qwen3 Coder 30B A3B Instruct | 13.6 |
| Qwen | Sep 2025 | Qwen3 Next 80B A3B Instruct | 13.8 |
| Qwen | Sep 2025 | Qwen3 Max | 24.5 |
| Qwen | Feb 2026 | Qwen3 Max Thinking | 32.5 |
| Qwen | Feb 2026 | Qwen3.5 397B A17B | 34.3 |
| Qwen | Feb 2026 | Qwen3.5-27B | 34.6 |
| Qwen | Apr 2026 | Qwen3.6 Plus | 40.5 |
| Qwen | Apr 2026 | Qwen3.6 Max Preview | 41.1 |
| Qwen | May 2026 | Qwen3.7 Max | 46.7 |
| Qwen | Aug 2026 | Qwen3.8 Max | 58.1 |
| DeepSeek | Jan 2025 | R1 | 20.4 |
| DeepSeek | Sep 2025 | V3.1 Terminus | 21.7 |
| DeepSeek | Dec 2025 | V3.2 | 25.1 |
| DeepSeek | Apr 2026 | V4 Flash 0423 | 51.8 |
| Meta | Apr 2025 | Llama 4 Scout | 10.3 |
| Meta | Apr 2025 | Llama 4 Maverick | 14.5 |
Cadence
Releases and standings
How many models shipped per quarter, and where each creator's best model currently sits.
Models released per quarter
Count of models entering the index · last twelve quarters
- 2023-Q35
- 2024-Q14
- 2024-Q25
- 2024-Q317
- 2024-Q414
- 2025-Q124
- 2025-Q228
- 2025-Q346
- 2025-Q444
- 2026-Q147
- 2026-Q256
- 2026-Q331
Counts the models tracked here, not every model released anywhere.
View as table
| Quarter | Models |
|---|---|
| 2023-Q3 | 5 |
| 2024-Q1 | 4 |
| 2024-Q2 | 5 |
| 2024-Q3 | 17 |
| 2024-Q4 | 14 |
| 2025-Q1 | 24 |
| 2025-Q2 | 28 |
| 2025-Q3 | 46 |
| 2025-Q4 | 44 |
| 2026-Q1 | 47 |
| 2026-Q2 | 56 |
| 2026-Q3 | 31 |
Best model per creator
Intelligence Index of each creator's strongest release
One entry per creator — this is a standings table, not a model ranking.
View as table
| Creator | Best model | Index |
|---|---|---|
| Anthropic | Claude Opus 5 | 63.1 |
| OpenAI | GPT-5.6 Sol | 60.9 |
| Moonshot AI | Kimi K3 | 59.7 |
| Qwen | Qwen3.8 Max | 58.1 |
| Meta | Muse Spark 1.2 | 56.8 |
| xAI | Grok 4.5 | 55.8 |
| Z.ai | GLM 5.2 | 52.6 |
| Gemini 3.5 Flash | 52.0 | |
| DeepSeek | V4 Flash 0731 | 51.8 |
| MiniMax | MiniMax M3 | 45.4 |
| Xiaomi | MiMo-V2.5-Pro | 42.9 |
| Thinking Machines | Inkling | 42.3 |