Claude Opus 4.8 benchmark ranking on every AI leaderboard
closed weights · #33 overall, podium score 3.3 · top 10 on 3 of 10 · released May 28, 2026
10 of 10 boards
| Rank | Board | Score | |||
|---|---|---|---|---|---|
| 9 |  MCP AtlasAgents and tools · as “claude-opus-4-8 (max)” | Agents and tools | 82.2% | 34 | |
| 9 |  Document ArenaVision · as “claude-opus-4-8-high” | Vision | 1,475 | 38 | |
| 10 |  Epoch ECIReasoning · as “Claude Opus 4.8” | Reasoning | 158.3 | 100 | |
| 12 | Search ArenaAgents and tools · as “claude-opus-4-8” | Agents and tools | 1,204 | 33 | |
| 16 | Vision ArenaVision · as “claude-opus-4-8-high” | Vision | 1,283 | 100 | |
| 17 |  ARC-AGI-2Reasoning · as “Claude Opus 4.8 (High)” | Reasoning | 72.1% | 78 | |
| 18 | Text ArenaChat · as “claude-opus-4-8-high” | Chat | 1,481 | 100 | |
| 23 | GPQA DiamondReasoning · as “claude-opus-4-8_max” | Reasoning | 91.0% | 100 | |
| 25 | WebDev ArenaCoding · as “claude-opus-4-8-high” | Coding | 1,555 | 100 | |
| 26 |  LiveBenchReasoning · as “claude-opus-4-8-max-effort” | Reasoning | 76.2% | 59 |
Category ranks and more models
Sources: each leaderboard's own published results page, captured every day at 04:30 UTC. Scores exactly as each board published them; we run no evaluations. Logos via logo.dev; trademarks belong to their owners.