B.O.B. Leaderboard
Anthropicreleased 16 Apr 2026
Claude Opus 4.7
#23 of 9534 of 46 benchmarksConsistent
62.4
B.O.B.
Category profile
- Real use (hard to game)2×66
- Reasoning & knowledge48
- Math59
- Coding68
- Agents & real work61
- Independent indexes69
Reality check
Lab tests
59
Real use (6 tests)
66
Lab tests and real-use tests agree. It does 7 points better on tests that are hard to train for than on the usual lab tests.
What counts as real usePrice per 1M tokens (OpenRouter)
$5.00 in · $25.00 out
Cost per task on the AA index
—
Output speed (Artificial Analysis)
—
§Receipts
Every benchmark result
“Beats” = the share of tracked models this one beat on that test. The raw score is its best reported setting.
| Benchmark | Raw | Beats | Source |
|---|---|---|---|
| Real use (hard to game) · Real use 66 | |||
| LMArena Textas claude-opus-4-7-high | 1501 | 94%of 80 | LMArena (blind human votes) |
| Kagi LLM Benchmarkas claude-opus-4-7-thinking | 80.7% | 76%of 22 | Kagi |
| EQ-Bench Creative Writingas claude-opus-4-7 | 1914 | 76%of 46 | Sam Paech (open source) |
| LMArena WebDevas claude-opus-4-7-high | 1558 | 61%of 82 | LMArena (blind human votes) |
| SWE-rebenchas Claude Opus 4.7-high | 53.1% | 46%of 34 | Nebius |
| SimpleBenchas claude-opus-4-7_unknown | 61.7% | 45%of 30 | Epoch AI Benchmarking Hub |
| Reasoning & knowledge · Reasoning 48 | |||
| ARC-AGI-2as claude-opus-4-7_max | 75.8% | 63%of 44 | Epoch AI Benchmarking Hub |
| SimpleQA Verifiedas claude-opus-4-7_xhigh | 51.7% | 61%of 57 | Epoch AI Benchmarking Hub |
| Chess Puzzlesas claude-opus-4-7_max | 41.0% | 59%of 73 | Epoch AI Benchmarking Hub |
| CritPt (physics research)as claude-opus-4-7_max | 12.0% | 51%of 89 | Epoch AI Benchmarking Hub |
| GPQA Diamondas claude-opus-4-7_xhigh | 90.2% | 50%of 70 | Epoch AI Benchmarking Hub |
| Mystery Game Puzzlesas claude-opus-4-7_max | 28.0% | 48%of 57 | Epoch AI Benchmarking Hub |
| ForecastBenchas claude-opus-4-7_unknown | 59.9% | 47%of 18 | Epoch AI Benchmarking Hub |
| Humanity's Last Examas claude-opus-4-7_unknown | 36.2% | 33%of 13 | Epoch AI Benchmarking Hub |
| Blueprint-Bench 2as claude-opus-4-7_unknown | 24.5% | 23%of 31 | Epoch AI Benchmarking Hub |
| Math · Math 59 | |||
| OTIS Mock AIMEas claude-opus-4-7_xhigh | 97.8% | 67%of 73 | Epoch AI Benchmarking Hub |
| FrontierMath Tiers 1-3as claude-opus-4-7_max | 70.2% | 61%of 60 | Epoch AI Benchmarking Hub |
| ProofBenchas claude-opus-4-7_max | 54.0% | 60%of 68 | Epoch AI Benchmarking Hub |
| FrontierMath Tier 4as claude-opus-4-7_max | 31.7% | 48%of 52 | Epoch AI Benchmarking Hub |
| Coding · Coding 68 | |||
| SWE-bench Verifiedas claude-opus-4-7_max | 83.5% | 100%of 16 | Epoch AI Benchmarking Hub |
| Terminal-Benchas claude-opus-4-7_unknown | 80.2% | 81%of 14 | Epoch AI Benchmarking Hub |
| WeirdMLas claude-opus-4-7_high | 76.4% | 71%of 52 | Epoch AI Benchmarking Hub |
| ALE-Benchas claude-opus-4-7 | 1323.0% | 69%of 71 | Epoch AI Benchmarking Hub |
| SciCodeas claude-opus-4-7_max | 54.5% | 66%of 86 | Epoch AI Benchmarking Hub |
| GSO (code optimisation)as claude-opus-4-7_high | 44.1% | 50%of 11 | Epoch AI Benchmarking Hub |
| FrontierCodeas claude-opus-4-7_unknown | 38.5% | 36%of 45 | Epoch AI Benchmarking Hub |
| Agents & real work · Agents 61 | |||
| Vending-Bench 2as claude-opus-4-7_unknown | 10936.8% | 89%of 47 | Epoch AI Benchmarking Hub |
| LMCAas claude-opus-4-7_max | 52.2% | 73%of 67 | Epoch AI Benchmarking Hub |
| DTBenchas claude-opus-4-7_max | 94.7% | 68%of 67 | Epoch AI Benchmarking Hub |
| OSWorld 2.0as claude-opus-4-7_max | 18.2% | 63%of 9 | Epoch AI Benchmarking Hub |
| Game Boy Advance evalas claude-opus-4-7_unknown | 43.8% | 50%of 23 | Epoch AI Benchmarking Hub |
| APEX-Agentsas claude-opus-4-7_max | 49.2% | 45%of 48 | Epoch AI Benchmarking Hub |
| GDP.pdfas claude-opus-4-7_max | 21.0% | 42%of 37 | Epoch AI Benchmarking Hub |
| Independent indexes · Indexes 69 | |||
| Epoch Capabilities Indexas Claude Opus 4.7 | 156.3 | 69%of 81 | Epoch AI |