History
LLM rankings over time
The top spot on the LMArena text leaderboard has changed hands 17 times between May 2023 and October 2026. Gemini 4 Argon has held it since September 2026.
Last verified
Best rating per lab, by month
Highest LMArena text rating held by any model from each lab. Ratings are relative, so later values mean beating stronger competition.
Every change at #1
| Month | New leader | Rating |
|---|---|---|
| September 2026 | Gemini 4 Argon | 1533 |
| July 2026 | Claude Opus 5 | 1512 |
| February 2026 | Claude Opus 4.6 | 1501 |
| November 2025 | Gemini 3 Pro | 1488 |
| June 2025 | Gemini 2.5 Pro | 1478 |
| April 2025 | o3 | 1418 |
| March 2025 | Chatgpt 4o Latest 20250326 | 1406 |
| January 2025 | DeepSeek-R1 | 1358 |
| September 2024 | o1 | 1339 |
| August 2024 | Chatgpt 4o Latest 20250326 | 1316 |
| May 2024 | GPT-4o | 1282 |
| April 2024 | GPT-4 Turbo | 1257 |
| March 2024 | Claude 3 Opus | 1254 |
| December 2023 | GPT-4 | 1129 |
| October 2023 | Wizardlm 70b | 1099 |
| September 2023 | Llama 2-70B | 1060 |
| June 2023 | Wizardlm 13b | 1048 |
| May 2023 | Palm 2 | 1059 |
Noometry Index snapshots
| Date | Top five |
|---|---|
| 2026-10-10 | GPT-6 Astra 70.8, Claude Fable 5.1 69.0, Claude Opus 5.5 68.6, Claude Opus 5 67.8, Claude Fable 5 66.8 |
| 2026-09-30 | GPT-6 Astra 70.3, Claude Opus 5.5 68.6, Claude Fable 5.1 68.5, Claude Opus 5 67.8, Claude Fable 5 66.4 |
| 2026-09-27 | GPT-6 Astra 64.6, Claude Fable 5.1 62.5, Claude Opus 5 61.2, Claude Fable 5 60.3, GPT-5.6 Sol 58.9 |
Frequently asked questions
Which AI model has been #1 the longest?
The top spot on the LMArena text leaderboard has changed hands 17 times between May 2023 and October 2026. Gemini 4 Argon has held it since September 2026.
Where does this history come from?
Monthly ratings from the public LMArena leaderboard dataset (blind human votes, style control on), plus daily snapshots of the Noometry Index since it launched.