WebDev Arena
Human preference for web development: people compare what two anonymous models build for the same prompt and vote. Scores are Bradley-Terry ratings with 95% confidence intervals.
- Unit
- Arena score
- Results
- 148 results110 Tokscale models matched, 129 in the source
- Last update
- Oct 19, 2026Newest publish date
- Retrieved
- Oct 21, 2026Source data through Oct 19, 2026
Conditions. The overall category of LMArena's webdev board (now called Code Arena). Reasoning effort is part of each entry's name.
Attribution. LMArena leaderboard dataset (lmarena-ai/leaderboard-dataset on Hugging Face), CC BY 4.0
Results
One row per model: its best reported result. The rank is the position in this list, by score; the source's own ranking can differ.
Tokscale usage counts accounts that used the model in public Tokscale submissions, Oct 14 – Nov 12, 2026. It is shown next to the score and does not explain it.
| Rank | Model | Score (Arena score) | Conditions | Date | Tokscale usage (30 days) | Source link |
|---|---|---|---|---|---|---|
| 1 | Claude Opus 5.5Listed as Claude Opus 5.5 (Max) | 1,807Confidence interval 1793–18212,925 votes | Max effort | Oct 19, 2026Published | 709active users | Source for Claude Opus 5.5 (opens in a new tab) |
| 2 | GPT-6 AstraListed as GPT-6 Astra (Max) | 1,774Confidence interval 1764–17846,169 votes | Max effort | Oct 19, 2026Published | 562active users | Source for GPT-6 Astra (opens in a new tab) |
| 3 | Claude Sonnet 5.5Listed as Claude Sonnet 5.5 (xHigh) | 1,768Confidence interval 1756–17813,428 votes | XHigh effort | Oct 19, 2026Published | 555active users | Source for Claude Sonnet 5.5 (opens in a new tab) |
| 4 | GPT-6.1 SolListed as GPT-6.1 Sol (Max) | 1,753Confidence interval 1740–17663,072 votes | Max effort | Oct 19, 2026Published | 659active users | Source for GPT-6.1 Sol (opens in a new tab) |
| 5 | Claude Fable 5.1Listed as Claude Fable 5.1 (Max) | 1,743Confidence interval 1734–17527,744 votes | Max effort | Oct 19, 2026Published | 361active users | Source for Claude Fable 5.1 (opens in a new tab) |
| 6 | Muse Wonder Full 1.0Listed as Muse Wonder Full (Max) | 1,726Confidence interval 1710–17421,639 votes | Max effort | Oct 19, 2026Published | 10active users | Source for Muse Wonder Full 1.0 (opens in a new tab) |
| 7 | Claude Opus 5Listed as Claude Opus 5 (Max) | 1,689Confidence interval 1683–169518,555 votes | Max effort | Oct 19, 2026Published | 163active users | Source for Claude Opus 5 (opens in a new tab) |
| 8 | GPT-6 SolListed as GPT-6 Sol (Max) | 1,683Confidence interval 1673–16934,608 votes | Max effort | Oct 19, 2026Published | 245active users | Source for GPT-6 Sol (opens in a new tab) |
| 9 | Gemini 4 ArgonListed as Gemini 4 Argon (High) | 1,678Confidence interval 1663–16941,728 votes | High effort | Oct 19, 2026Published | 82active users | Source for Gemini 4 Argon (opens in a new tab) |
| 10 | Qwen3.8 Max 0902Listed as Qwen3.8-Max (0902) | 1,674Confidence interval 1667–168114,421 votes | Not reported | Oct 19, 2026Published | 1active users | Source for Qwen3.8 Max 0902 (opens in a new tab) |
| 11 | Qwen3.8 MaxListed as Qwen3.8-Max | 1,672Confidence interval 1660–16843,428 votes | Not reported | Oct 19, 2026Published | 16active users | Source for Qwen3.8 Max (opens in a new tab) |
| 12 | DeepSeek V4.1 FlashListed as DeepSeek-V4.1-Flash-20261017 (Max) | 1,664Confidence interval 1650–16792,008 votes | Max effort | Oct 19, 2026Published | 225active users | Source for DeepSeek V4.1 Flash (opens in a new tab) |
| 13 | Muse Spark 1.3Listed as Muse Spark 1.3 (Max) | 1,659Confidence interval 1651–16679,407 votes | Max effort | Oct 19, 2026Published | 21active users | Source for Muse Spark 1.3 (opens in a new tab) |
| 14 | Kimi K3Listed as Kimi K3 (Max) | 1,655Confidence interval 1649–166119,592 votes | Max effort | Oct 19, 2026Published | 56active users | Source for Kimi K3 (opens in a new tab) |
| 15 | Muse Wonder 1.0Listed as Muse Wonder (Max) | 1,652Confidence interval 1638–16652,289 votes | Max effort | Oct 19, 2026Published | 17active users | Source for Muse Wonder 1.0 (opens in a new tab) |
| 16 | Grok 4.7Listed as Grok 4.7 (xHigh) | 1,637Confidence interval 1627–16475,303 votes | XHigh effort | Oct 19, 2026Published | 131active users | Source for Grok 4.7 (opens in a new tab) |
| 17 | MiMo V2.6 FlashListed as MiMo-V2.6-Flash | 1,634Confidence interval 1624–16435,293 votes | Not reported | Oct 19, 2026Published | 23active users | Source for MiMo V2.6 Flash (opens in a new tab) |
| 18 | Hy4 PreviewListed as Hy4 preview | 1,631Confidence interval 1621–16414,798 votes | Not reported | Oct 19, 2026Published | 21active users | Source for Hy4 Preview (opens in a new tab) |
| 19 | Qwen3.8 Flash NextListed as Qwen3.8-Flash-Next | 1,629Confidence interval 1621–16378,382 votes | Not reported | Oct 19, 2026Published | 6active users | Source for Qwen3.8 Flash Next (opens in a new tab) |
| 20 | MiMo V2.6 ProListed as MiMo-V2.6-Pro | 1,625Confidence interval 1615–16355,041 votes | Not reported | Oct 19, 2026Published | 39active users | Source for MiMo V2.6 Pro (opens in a new tab) |
| 21 | Claude Fable 5Listed as Claude Fable 5 (High) | 1,624Confidence interval 1618–163015,899 votes | High effort | Oct 19, 2026Published | 41active users | Source for Claude Fable 5 (opens in a new tab) |
| 22 | GLM-5.3Listed as GLM-5.3 (Max) | 1,621Confidence interval 1614–16299,828 votes | Max effort | Oct 19, 2026Published | 138active users | Source for GLM-5.3 (opens in a new tab) |
| 23 | Grok 4.6Listed as Grok 4.6 (High) | 1,617Confidence interval 1609–162410,483 votes | High effort | Oct 19, 2026Published | 41active users | Source for Grok 4.6 (opens in a new tab) |
| 24 | GPT-5.6 SolListed as GPT-5.6 Sol (xHigh, Codex Harness) | 1,616Confidence interval 1610–162217,786 votes | XHigh effort · Codex | Oct 19, 2026Published | 217active users | Source for GPT-5.6 Sol (opens in a new tab) |
| 25 | GLM-5.3 FlashListed as GLM-5.3-Flash | 1,607Confidence interval 1600–161318,264 votes | Not reported | Oct 19, 2026Published | 170active users | Source for GLM-5.3 Flash (opens in a new tab) |
| 26 | GLM-5.2Listed as GLM-5.2 (Max) | 1,604Confidence interval 1598–161016,657 votes | Max effort | Oct 19, 2026Published | 13active users | Source for GLM-5.2 (opens in a new tab) |
| 27 | Qwen3.8 27BListed as Qwen3.8-27B | 1,593Confidence interval 1586–159916,385 votes | Not reported | Oct 19, 2026Published | 6active users | Source for Qwen3.8 27B (opens in a new tab) |
| 28 | Gemini 3.7 FlashListed as Gemini 3.7 Flash (High) | 1,591Confidence interval 1584–159812,636 votes | High effort | Oct 19, 2026Published | 4active users | Source for Gemini 3.7 Flash (opens in a new tab) |
| 29 | MiniMax M3.1 FlashListed as MiniMax-M3.1-Flash | 1,587Confidence interval 1573–16021,853 votes | Not reported | Oct 19, 2026Published | 38active users | Source for MiniMax M3.1 Flash (opens in a new tab) |
| 30 | Claude Haiku 5.5Listed as Claude Haiku 5.5 (High) | 1,586Confidence interval 1575–15973,514 votes | High effort | Oct 19, 2026Published | 477active users | Source for Claude Haiku 5.5 (opens in a new tab) |
| 31 | Gemini 3.8 FlashListed as Gemini 3.8 Flash (High) | 1,585Confidence interval 1577–159210,851 votes | High effort | Oct 19, 2026Published | 66active users | Source for Gemini 3.8 Flash (opens in a new tab) |
| 32 | DeepSeek V4 ProListed as DeepSeek-V4-Pro (High, 20260813) | 1,582Confidence interval 1573–15924,714 votes | High effort | Oct 19, 2026Published | 51active users | Source for DeepSeek V4 Pro (opens in a new tab) |
| 33 | DeepSeek V4 FlashListed as DeepSeek-V4-Flash (High) | 1,582Confidence interval 1573–15914,947 votes | High effort | Oct 19, 2026Published | 51active users | Source for DeepSeek V4 Flash (opens in a new tab) |
| 34 | GPT-6 LunaListed as GPT-6 Luna (Max) | 1,578Confidence interval 1572–158421,622 votes | Max effort | Oct 19, 2026Published | 432active users | Source for GPT-6 Luna (opens in a new tab) |
| 35 | Step 5 PreviewListed as Step 5 Preview (High) | 1,567Confidence interval 1557–15774,728 votes | High effort | Oct 19, 2026Published | 30active users | Source for Step 5 Preview (opens in a new tab) |
| 36 | Claude Opus 4.7Listed as Claude Opus 4.7 (High) | 1,558Confidence interval 1552–156417,910 votes | High effort | Oct 19, 2026Published | 27active users | Source for Claude Opus 4.7 (opens in a new tab) |
| 37 | Claude Opus 4.8Listed as Claude Opus 4.8 (High) | 1,556Confidence interval 1550–156218,493 votes | High effort | Oct 19, 2026Published | 222active users | Source for Claude Opus 4.8 (opens in a new tab) |
| 38 | 1,553Confidence interval 1546–156011,082 votes | Not reported | Oct 19, 2026Published | 9active users | Source for Grok 4.5 (opens in a new tab) | |
| 39 | Claude Opus 4.6Listed as Claude Opus 4.6 (High) | 1,547Confidence interval 1541–155219,383 votes | High effort | Oct 19, 2026Published | 52active users | Source for Claude Opus 4.6 (opens in a new tab) |
| 40 | 1,543Confidence interval 1530–15552,832 votes | Not reported | Oct 19, 2026Published | 4active users | Source for Mistral Large 4 (opens in a new tab) | |
| 41 | 1,542Confidence interval 1535–15507,985 votes | Not reported | Oct 19, 2026Published | No Tokscale usage in the last 30 days | Source for Muse Spark 1.1 (opens in a new tab) | |
| 42 | Claude Sonnet 5Listed as Claude Sonnet 5 (High) | 1,540Confidence interval 1534–154615,755 votes | High effort | Oct 19, 2026Published | 138active users | Source for Claude Sonnet 5 (opens in a new tab) |
| 43 | Gemini 3.6 FlashListed as Gemini 3.6 Flash (High) | 1,538Confidence interval 1531–15468,762 votes | High effort | Oct 19, 2026Published | 3active users | Source for Gemini 3.6 Flash (opens in a new tab) |
| 44 | Muse Spark 1.2Listed as Muse Spark 1.2 (xHigh) | 1,534Confidence interval 1520–15472,163 votes | XHigh effort | Oct 19, 2026Published | No Tokscale usage in the last 30 days | Source for Muse Spark 1.2 (opens in a new tab) |
| 45 | Muse Spark OpenListed as Muse Spark Open (Max) | 1,528Confidence interval 1511–15451,380 votes | Max effort | Oct 19, 2026Published | 2active users | Source for Muse Spark Open (opens in a new tab) |
| 46 | 1,523Confidence interval 1518–152822,822 votes | Not reported | Oct 19, 2026Published | 55active users | Source for Claude Sonnet 4.6 (opens in a new tab) | |
| 47 | GPT-5.6 TerraListed as GPT-5.6 Terra (xHigh, Codex Harness) | 1,522Confidence interval 1515–152813,919 votes | XHigh effort · Codex | Oct 19, 2026Published | 119active users | Source for GPT-5.6 Terra (opens in a new tab) |
| 48 | Seed2.1 Pro PreviewNot matched to a Tokscale model | 1,520Confidence interval 1514–152615,123 votes | Not reported | Oct 19, 2026Published | Not matched to a Tokscale model | Source for Seed2.1 Pro Preview (opens in a new tab) |
| 49 | GPT-5.6 LunaListed as GPT-5.6 Luna (xHigh, Codex Harness) | 1,518Confidence interval 1512–152512,501 votes | XHigh effort · Codex | Oct 19, 2026Published | 158active users | Source for GPT-5.6 Luna (opens in a new tab) |
| 50 | Qwen3.7 MaxListed as Qwen3.7-Max (20260517) | 1,515Confidence interval 1508–15238,887 votes | Not reported | Oct 19, 2026Published | 3active users | Source for Qwen3.7 Max (opens in a new tab) |
Showing 50 of 127 models
Score by release date
Each dot is one model's best result on WebDev Arena, placed at the model's release date. The dashed line is the best score among models released up to that date.
- Used on Tokscale in the last 30 days
- No Tokscale usage in the last 30 days
- Best score so far
Claude Opus 5.5: 1,807 · released Sep 22, 2026 (date from models.dev) · Max effort · 709 active users on Tokscale in the last 30 days
100 models, released Jun 17, 2025 to Oct 19, 2026. Highest: Claude Opus 5.5, 1,807, released Sep 22, 2026 (date from models.dev). Use the left and right arrow keys to read each model in release order.1,0001,2001,4001,6001,8002,000
Release dates are from models.dev, because LMArena does not list them for this benchmark. 10 models have no known release date and are left out.
Show data
| Model | Released | Score | Tokscale users, 30 days |
|---|---|---|---|
| MiniMax M3.1 Flash | Oct 19, 2026models.dev | 1,587 | 38 |
| Muse Spark Open | Oct 15, 2026models.dev | 1,528Max effort | 2 |
| Gemini 4 Argon | Oct 13, 2026models.dev | 1,678High effort | 82 |
| Muse Wonder 1.0 | Oct 12, 2026models.dev | 1,652Max effort | 17 |
| Claude Haiku 5.5 | Oct 7, 2026models.dev | 1,586High effort | 477 |
| Mistral Large 4 | Oct 6, 2026models.dev | 1,543 | 4 |
| Beam 1.0 501B A23B | Oct 5, 2026models.dev | 1,437 | 1 |
| GPT-6.1 Sol | Sep 29, 2026models.dev | 1,753Max effort | 659 |
| Claude Sonnet 5.5 | Sep 28, 2026models.dev | 1,768XHigh effort | 555 |
| GPT-6 Luna | Sep 22, 2026models.dev | 1,578Max effort | 432 |
| MiMo V2.6 Pro | Sep 22, 2026models.dev | 1,625 | 39 |
| MiMo V2.6 Flash | Sep 22, 2026models.dev | 1,634 | 23 |
| GPT-6 Sol | Sep 22, 2026models.dev | 1,683Max effort | 245 |
| Claude Opus 5.5 | Sep 22, 2026models.dev | 1,807Max effort | 709 |
| Grok 4.7 | Sep 21, 2026models.dev | 1,637XHigh effort | 131 |
| Step 5 Preview | Sep 16, 2026models.dev | 1,567High effort | 30 |
| DeepSeek V4.1 Flash | Sep 10, 2026models.dev | 1,664Max effort | 225 |
| GPT-6 Astra | Sep 4, 2026models.dev | 1,774Max effort | 562 |
| Gemini 3.8 Flash | Sep 2, 2026models.dev | 1,585High effort | 66 |
| Muse Spark 1.3 | Sep 2, 2026models.dev | 1,659Max effort | 21 |
| Qwen3.8 Max 0902 | Sep 2, 2026models.dev | 1,674 | 1 |
| Claude Fable 5.1 | Sep 1, 2026models.dev | 1,743Max effort | 361 |
| Hy4 Preview | Aug 28, 2026models.dev | 1,631 | 21 |
| GLM-5.3 Flash | Aug 26, 2026models.dev | 1,607 | 170 |
| Qwen3.8 27B | Aug 14, 2026models.dev | 1,593 | 6 |
| GLM-5.3 | Aug 14, 2026models.dev | 1,621Max effort | 138 |
| Gemini 3.7 Flash | Aug 13, 2026models.dev | 1,591High effort | 4 |
| Grok 4.6 | Aug 12, 2026models.dev | 1,617High effort | 41 |
| Muse Spark 1.2 | Aug 5, 2026models.dev | 1,534XHigh effort | No Tokscale usage in the last 30 days |
| Qwen3.8 Max | Aug 3, 2026models.dev | 1,672 | 16 |
| Inkling Small | Jul 30, 2026models.dev | 1,409 | No Tokscale usage in the last 30 days |
| Claude Opus 5 | Jul 24, 2026models.dev | 1,689Max effort | 163 |
| Gemini 3.5 Flash Lite | Jul 21, 2026models.dev | 1,435 | 13 |
| Gemini 3.6 Flash | Jul 21, 2026models.dev | 1,538High effort | 3 |
| Kimi K3 | Jul 16, 2026models.dev | 1,655Max effort | 56 |
| Inkling | Jul 15, 2026models.dev | 1,414 | No Tokscale usage in the last 30 days |
| GPT-5.6 Luna | Jul 9, 2026models.dev | 1,518XHigh effort · Codex | 158 |
| GPT-5.6 Terra | Jul 9, 2026models.dev | 1,522XHigh effort · Codex | 119 |
| GPT-5.6 Sol | Jul 9, 2026models.dev | 1,616XHigh effort · Codex | 217 |
| Grok 4.5 | Jul 8, 2026models.dev | 1,553 | 9 |
| Hy3 | Jul 6, 2026models.dev | 1,507 | 18 |
| Claude Sonnet 5 | Jun 29, 2026models.dev | 1,540High effort | 138 |
| GLM-5.2 | Jun 13, 2026models.dev | 1,604Max effort | 13 |
| Kimi K2.7 Code | Jun 12, 2026models.dev | 1,474 | 17 |
| Claude Fable 5 | Jun 7, 2026models.dev | 1,624High effort | 41 |
| MiniMax M3 | Jun 1, 2026models.dev | 1,482 | 26 |
| Claude Opus 4.8 | May 28, 2026models.dev | 1,556High effort | 222 |
| Qwen3.7 Max | May 21, 2026models.dev | 1,515 | 3 |
| Gemini 3.5 Flash | May 19, 2026models.dev | 1,499High effort | 3 |
| DeepSeek V4 Flash | Apr 24, 2026models.dev | 1,582High effort | 51 |
| DeepSeek V4 Pro | Apr 24, 2026models.dev | 1,582High effort | 51 |
| GPT-5.5 | Apr 23, 2026models.dev | 1,513XHigh effort · Codex | 30 |
| MiMo V2.5 | Apr 22, 2026models.dev | 1,438 | 5 |
| MiMo V2.5 Pro | Apr 22, 2026models.dev | 1,480 | 3 |
| Hy3 Preview | Apr 20, 2026models.dev | 1,357 | No Tokscale usage in the last 30 days |
| Qwen3.6 Max Preview | Apr 20, 2026models.dev | 1,482 | No Tokscale usage in the last 30 days |
| Kimi K2.6 | Apr 20, 2026models.dev | 1,509 | 4 |
| Grok 4.3 | Apr 17, 2026models.dev | 1,358 | No Tokscale usage in the last 30 days |
| Claude Opus 4.7 | Apr 14, 2026models.dev | 1,558High effort | 27 |
| Muse Spark 1.1 | Apr 8, 2026models.dev | 1,542 | No Tokscale usage in the last 30 days |
| GLM-5.1 | Apr 7, 2026models.dev | 1,508 | 5 |
| Qwen3.6 Plus | Apr 2, 2026models.dev | 1,461 | 1 |
| Trinity Large Thinking | Apr 1, 2026models.dev | 1,238 | 1 |
| GLM-5v Turbo | Apr 1, 2026models.dev | 1,402 | No Tokscale usage in the last 30 days |
| MiniMax M2.7 | Mar 18, 2026models.dev | 1,399 | 6 |
| MiMo V2 Pro | Mar 18, 2026models.dev | 1,434 | No Tokscale usage in the last 30 days |
| GPT-5.4 Mini | Mar 17, 2026models.dev | 1,397High effort | 1 |
| GPT-5.4 | Mar 5, 2026models.dev | 1,464High effort · Codex | 1 |
| Gemini 3.1 Flash Lite Preview | Mar 3, 2026models.dev | 1,257 | 1 |
| Mercury 2 | Feb 24, 2026models.dev | 1,171 | 1 |
| Qwen3.5 Flash | Feb 23, 2026models.dev | 1,244 | No Tokscale usage in the last 30 days |
| Qwen3.5 35B A3B | Feb 23, 2026models.dev | 1,254 | No Tokscale usage in the last 30 days |
| Qwen3.5 27B | Feb 23, 2026models.dev | 1,358 | No Tokscale usage in the last 30 days |
| Qwen3.5 122B A10B | Feb 23, 2026models.dev | 1,361 | No Tokscale usage in the last 30 days |
| Gemini 3.1 Pro Preview | Feb 19, 2026models.dev | 1,447 | 5 |
| Claude Sonnet 4.6 | Feb 17, 2026models.dev | 1,523 | 55 |
| Qwen3.5 397B A17B | Feb 15, 2026models.dev | 1,400 | No Tokscale usage in the last 30 days |
| MiniMax M2.5 | Feb 12, 2026models.dev | 1,386 | 3 |
| GLM-5 | Feb 12, 2026models.dev | 1,435 | 2 |
| GPT-5.3 Codex | Feb 5, 2026models.dev | 1,407Codex | 5 |
| Claude Opus 4.6 | Feb 4, 2026models.dev | 1,547High effort | 52 |
| Kimi K2.5 | Jan 27, 2026models.dev | 1,437 | No Tokscale usage in the last 30 days |
| GLM-4.7 | Dec 22, 2025models.dev | 1,435 | 4 |
| MiMo V2 Flash | Dec 16, 2025models.dev | 1,330 | No Tokscale usage in the last 30 days |
| GPT-5.2 Codex | Dec 11, 2025models.dev | 1,339 | No Tokscale usage in the last 30 days |
| GPT-5.2 | Dec 11, 2025models.dev | 1,417 | No Tokscale usage in the last 30 days |
| DeepSeek V3.2 | Dec 1, 2025models.dev | 1,362 | 1 |
| Claude Opus 4.5 | Nov 24, 2025models.dev | 1,494High effort · 32K thinking budget | 4 |
| Grok 4.1 Fast | Nov 19, 2025models.dev | 1,242 | No Tokscale usage in the last 30 days |
| GPT-5.1 Codex Mini | Nov 13, 2025models.dev | 1,244 | No Tokscale usage in the last 30 days |
| GPT-5.1 Codex | Nov 13, 2025models.dev | 1,337 | No Tokscale usage in the last 30 days |
| GPT-5.1 | Nov 13, 2025models.dev | 1,396Medium effort | No Tokscale usage in the last 30 days |
| MiniMax M2 | Oct 27, 2025models.dev | 1,297 | No Tokscale usage in the last 30 days |
| Claude Haiku 4.5 | Oct 15, 2025models.dev | 1,330 | 102 |
| GLM-4.6 | Sep 30, 2025models.dev | 1,341 | 2 |
| Claude Sonnet 4.5 | Sep 29, 2025models.dev | 1,394High effort · 32K thinking budget | 23 |
| GPT-5 | Aug 7, 2025models.dev | 1,418Medium effort | No Tokscale usage in the last 30 days |
| Claude Opus 4.1 | Aug 5, 2025models.dev | 1,390 | No Tokscale usage in the last 30 days |
| Devstral Medium 2507 | Jul 10, 2025models.dev | 1,071 | No Tokscale usage in the last 30 days |
| Gemini 2.5 Pro | Jun 17, 2025models.dev | 1,226 | 1 |
Method and matching
How results are run
- The overall category of LMArena's webdev board (now called Code Arena). Reasoning effort is part of each entry's name.
- Ratings from different arenas are on different scales and cannot be compared with each other.
Matching to Tokscale models
- 131 of 148 results are matched to 110 Tokscale models: 54 by exact id, 75 by normalized name, 2 by a manual rule.
- No match is marked as reviewed by hand yet.
- 17 results from 17 models are not matched. They stay in the table without a Tokscale link.
- Reasoning effort, thinking budget and agent are conditions of a result. They never change which model a result belongs to.
License and attribution
- CC BY 4.0 (opens in a new tab), retrieved Oct 21, 2026.
- LMArena leaderboard dataset (lmarena-ai/leaderboard-dataset on Hugging Face), CC BY 4.0
Sources and attribution
Each score keeps its own source, unit and run conditions. Tokscale does not combine benchmarks into one score, and a missing score means the model was not evaluated, not that it scored zero.