B.O.B. Leaderboard
OpenAIreleased 9 Jul 2026
GPT-5.6 Terra
#20 of 9533 of 46 benchmarksBenchmaxxed?
63.6
B.O.B.
Category profile
- Real use (hard to game)2×42
- Reasoning & knowledge67
- Math80
- Coding66
- Agents & real work70
- Independent indexes78
Reality check
Lab tests
71
Real use (9 tests)
42
Scores much higher on lab tests than on real-use tests. It does 28 points better on the usual lab tests than on tests that are hard to train for.
What counts as real usePrice per 1M tokens (OpenRouter)
$2.00 in · $12.00 out
Cost per task on the AA index
$1.40
Output speed (Artificial Analysis)
127 tok/s
§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 42 | |||
| EQ-Bench Creative Writingas gpt-5.6-terra | 1855 | 71%of 46 | Sam Paech (open source) |
| LMArena Textas gpt-5.6-terra-xhigh | 1466 | 56%of 80 | LMArena (blind human votes) |
| Vals Indexas openai/gpt-5.6-terra | 53.1% | 54%of 42 | Vals AI (private test sets) |
| LiveBench 2.0 (real-work suites)as gpt-5.6-terra | 71.0% | 51%of 39 | LiveBench / Abacus.AI |
| LMArena WebDevas gpt-5.6-terra-xhigh (codex-harness) | 1522 | 48%of 82 | LMArena (blind human votes) |
| LMArena Agent Arenaas GPT 5.6 Terra (xHigh) | 0.005 | 43%of 50 | LMArena (real agent sessions) |
| BridgeBench (overall)as GPT-5.6 Terra | 511 | 25%of 29 | BridgeMind |
| Kagi LLM Benchmarkas gpt-5-6-terra | 51.3% | 19%of 22 | Kagi |
| SimpleBenchas gpt-5.6-terra_unknown | 48.9% | 14%of 30 | Epoch AI Benchmarking Hub |
| Reasoning & knowledge · Reasoning 67 | |||
| CritPt (physics research)as gpt-5.6-terra_max | 30.0% | 90%of 89 | Epoch AI Benchmarking Hub |
| GPQA Diamondas gpt-5.6-terra_max | 93.3% | 80%of 70 | Epoch AI Benchmarking Hub |
| Chess Puzzlesas gpt-5.6-terra_max | 47.0% | 74%of 73 | Epoch AI Benchmarking Hub |
| ARC-AGI-2as gpt-5.6-terra_max | 83.9% | 70%of 44 | Epoch AI Benchmarking Hub |
| Mystery Game Puzzlesas gpt-5.6-terra_max | 35.0% | 67%of 57 | Epoch AI Benchmarking Hub |
| Blueprint-Bench 2as gpt-5.6-terra_unknown | 30.8% | 47%of 31 | Epoch AI Benchmarking Hub |
| SimpleQA Verifiedas gpt-5.6-terra_max | 43.2% | 43%of 57 | Epoch AI Benchmarking Hub |
| Math · Math 80 | |||
| OTIS Mock AIMEas gpt-5.6-terra_max | 99.7% | 86%of 73 | Epoch AI Benchmarking Hub |
| FrontierMath Tiers 1-3as gpt-5.6-terra_max | 86.0% | 84%of 60 | Epoch AI Benchmarking Hub |
| ProofBenchas gpt-5.6-terra_xhigh | 74.0% | 78%of 68 | Epoch AI Benchmarking Hub |
| FrontierMath Tier 4as gpt-5.6-terra_max | 70.7% | 73%of 52 | Epoch AI Benchmarking Hub |
| Coding · Coding 66 | |||
| ALE-Benchas gpt-5.6-terra_max | 1951.4% | 89%of 71 | Epoch AI Benchmarking Hub |
| DeepSWEas gpt-5.6-terra_max | 69.6% | 80%of 26 | Epoch AI Benchmarking Hub |
| WeirdMLas gpt-5.6-terra_high | 78.3% | 77%of 52 | Epoch AI Benchmarking Hub |
| SciCodeas gpt-5.6-terra_max | 55.0% | 69%of 86 | Epoch AI Benchmarking Hub |
| FrontierCodeas gpt-5.6-terra_unknown | 41.3% | 48%of 45 | Epoch AI Benchmarking Hub |
| CursorBenchas gpt-5.6-terra_max | 41.3% | 31%of 17 | Epoch AI Benchmarking Hub |
| Agents & real work · Agents 70 | |||
| LMCAas gpt-5.6-terra_max | 55.0% | 83%of 67 | Epoch AI Benchmarking Hub |
| APEX-Agentsas gpt-5.6-terra_max | 58.2% | 72%of 48 | Epoch AI Benchmarking Hub |
| Vending-Bench 2as gpt-5.6-terra_unknown | 7343.2% | 67%of 47 | Epoch AI Benchmarking Hub |
| GDP.pdfas gpt-5.6-terra_unknown | 24.7% | 67%of 37 | Epoch AI Benchmarking Hub |
| DTBenchas gpt-5.6-terra_max | 93.3% | 61%of 67 | Epoch AI Benchmarking Hub |
| Independent indexes · Indexes 78 | |||
| Epoch Capabilities Indexas GPT-5.6 Terra | 159.4 | 86%of 81 | Epoch AI |
| Artificial Analysis Intelligence Indexas GPT-5.6 Terra (max) | 42 | 69%of 54 | Artificial Analysis |