gpt-5.6-sol
OpenAI- average±1.8 (95%)
- 86.8
- leaderboard rankof 114
- ≈8
- benchmarksof 12
- 12
- per 1,000 documentsmean over its benchmarks
- $62
- seconds per inputmean response time
- 27.5
By task
-
Data correction
76.2
Best of 114 models with all its benchmarks: claude-opus-5-5, 82.7.
-
Document understanding
96.0
Best of 107 models with all its benchmarks: gpt-6-astra, 97.0.
-
Information extraction
87.4
Best of 107 models with all its benchmarks: grok-4.6, 90.8.
-
Transcription
93.0
Best of 114 models with all its benchmarks: gemini-4-argon, 96.2.
By benchmark
| Benchmark | ScoreThe best current run on each benchmark, on the benchmark's metric scaled to 0–100. ± is the 95% sampling margin, from resampling the run's inputs. | RankRank among all current models on that benchmark; ≈ for models that cannot be told apart within sampling error. | Per 1,000 docs | s / input | Best run |
|---|---|---|---|---|---|
| Bibliographic Data | 97.1±1.1 | ≈29of 115 | $67$0.33 run | 37.59 | T1345 · 16 Sep 2026prompt.txt |
| Blacklist Cards | 96.4±2.0 | ≈14of 114 | $23$0.77 run | 7.65 | T1352 · 25 Jul 2026prompt.txt |
| Book Advert XML Repair | 54.6±10.1 | ≈1of 119 | $20$1.00 run | 15.57 | T1355 · 16 Sep 2026prompt.txt |
| Business Letters | 63.0±7.4 | ≈24of 114 | $31$1.75 run | 7.0 | T1346 · 16 Sep 2026prompt.txt |
| Company Lists | 86.3±13.0 | ≈29of 114 | $36$0.53 run | 14.06 | T1353 · 16 Sep 2026prompt.txt |
| Duty Rosters | 93.5±1.6 | ≈5of 107 | $110$2.30 run | 42.34 | T1359 · 25 Jul 2026prompt.txt |
| Fraktur Adverts | 97.7±0.5 | ≈15of 114 | $176$0.88 run | 88.74 | T1349 · 25 Jul 2026prompt_optimized.txt |
| General Meeting Minutes | 88.6±7.3 | ≈1of 108 | $64$0.58 run | 21.19 | T1358 · 25 Jul 2026prompt.txt |
| Library Cards | 86.9±1.4 | ≈41of 114 | $19$4.94 run | 6.63 | T1350 · 25 Jul 2026prompt.txt |
| Magazine Pages | 96.6±2.8 | ≈1of 115 | $21$0.96 run | 8.56 | T1357 · 16 Sep 2026prompt.txt |
| Medieval Manuscripts | 83.5±7.3 | ≈1of 114 | $127$1.52 run | 67.74 | T1351 · 25 Jul 2026prompt.txt |
| Personnel Cards | 97.8±0.9 | ≈1of 114 | $55$3.35 run | 13.43 | T1356 · 25 Jul 2026prompt.txt |
Price
| From | Input, $ per million tokens | Output, $ per million tokens | Source |
|---|---|---|---|
| 16 Sep 2026 | 4.0 | 20.0 | archived price page |
| 25 Jul 2026 | 5 | 30 | archived price page |
Every run
- T1359Duty Rosters16 Sep 202692.8$1.66
- T1358General Meeting Minutes16 Sep 202686.4$0.46
- T1357Magazine Pages16 Sep 202696.6$0.96
- T1356Personnel Cards16 Sep 202697.5$2.51
- T1355Book Advert XML Repair16 Sep 202654.6$1.00
- T1354Company Lists16 Sep 202670.8$0.52
- T1353Company Lists16 Sep 202686.3$0.53
- T1352Blacklist Cards16 Sep 202694.8$0.60
- T1351Medieval Manuscripts16 Sep 202678.0$1.15
- T1350Library Cards16 Sep 202686.7$3.81
- T1349Fraktur Adverts16 Sep 202695.3$0.51
- T1348Business Letters16 Sep 2026 · 32 signed letters63.0$1.09
- T1347Business Letters16 Sep 2026 · 25 unsigned letters73.0$0.62
- T1346Business Letters16 Sep 2026 · all 57 letters63.0$1.75
- T1345Bibliographic Data16 Sep 202697.1$0.33
- T1359Duty Rosters25 Jul 202693.5$2.30
- T1358General Meeting Minutes25 Jul 202688.6$0.58
- T1357Magazine Pages25 Jul 202694.7$1.22
- T1356Personnel Cards25 Jul 202697.8$3.35
- T1355Book Advert XML Repair25 Jul 202652.9$1.36
- T1354Company Lists25 Jul 202672.7$0.73
- T1353Company Lists25 Jul 202682.5$0.79
- T1352Blacklist Cards25 Jul 202696.4$0.77
- T1351Medieval Manuscripts25 Jul 202683.5$1.52
- T1350Library Cards25 Jul 202686.9$4.94
- T1349Fraktur Adverts25 Jul 202697.7$0.88
- T1348Business Letters25 Jul 2026 · 32 signed letters58.0$1.33
- T1347Business Letters25 Jul 2026 · 25 unsigned letters62.0$0.75
- T1346Business Letters25 Jul 2026 · all 57 letters59.0$2.12
- T1345Bibliographic Data25 Jul 202696.0$0.46