gpt-4.1
OpenAI- average
- 67.9
- leaderboard rankof 114
- ≈70
- benchmarksof 12
- 12
- per 1,000 documentsmean over its benchmarks
- $10
- seconds per inputmean response time
- 36.8
By task
-
Data correction
56.0
Best of 114 models with all its benchmarks: claude-opus-5-5, 82.7.
-
Document understanding
58.2
Best of 107 models with all its benchmarks: gpt-6-astra, 97.0.
-
Information extraction
79.1
Best of 107 models with all its benchmarks: grok-4.6, 90.8.
-
Transcription
76.5
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 Dataolder run | 95.1 | ≈55of 115 | $19$0.095 run | 59.79 | T0139 · 1 Oct 2025prompt.txt |
| Blacklist Cards | 93.0±3.1 | ≈51of 114 | $4.07$0.13 run | 6.29 | T0232 · 26 Jan 2026prompt.txt |
| Book Advert XML Repair | 20.9±6.1 | ≈64of 119 | $11$0.56 run | 26.0 | T0448 · 25 Jan 2026prompt.txt |
| Business Letters | 53.0±6.6 | ≈90of 114 | $5.30$0.30 run | 4.71 | T0067 · 24 Jan 2026prompt.txt |
| Company Lists | 72.5±12.3 | ≈80of 114 | $7.89$0.12 run | 17.78 | T0341 · 25 Jan 2026prompt.txt |
| Duty Rosters | 73.5±2.9 | ≈64of 107 | $22$0.47 run | 11.86 | T1293 · 16 Sep 2026prompt.txt |
| Fraktur Adverts | 71.7±15.6 | ≈67of 114 | $15$0.075 run | 281.43 | T0083 · 24 Jan 2026prompt_optimized.txt |
| General Meeting Minutes | 77.4±11.4 | ≈25of 108 | $8.44$0.076 run | 9.11 | T1244 · 3 Jul 2026prompt.txt |
| Library Cardsolder run | 89.4 | ≈15of 114 | –– run | 7.16 | T0160 · 2 Sep 2025prompt.txt |
| Magazine Pages | 9.8±8.4 | ≈66of 115 | $2.91$0.13 run | 3.36 | T0772 · 23 Mar 2026prompt.txt |
| Medieval Manuscripts | 66.8±13.1 | ≈72of 114 | $5.16$0.062 run | 9.59 | T0273 · 26 Jan 2026prompt.txt |
| Personnel Cards | 91.2±3.5 | ≈78of 114 | $12$0.70 run | 5.09 | T0605 · 16 Sep 2026prompt.txt |
Price
| From | Input, $ per million tokens | Output, $ per million tokens | Source |
|---|---|---|---|
| 0 Sep 2025 | 2.0 | 8.0 | archived price page |
Every run
- T1293Duty Rosters16 Sep 202673.5$0.47
- T1244General Meeting Minutes16 Sep 202674.5$0.076
- T0772Magazine Pages16 Sep 20267.6$0.13
- T0605Personnel Cards16 Sep 202691.2$0.70
- T1293Duty Rosters24 Jul 202671.9$0.46
- T1244General Meeting Minutes3 Jul 202677.4$0.076
- T0772Magazine Pages23 Mar 20269.8$0.13
- T0605Personnel Cards9 Feb 202690.1$0.71
- T0273Medieval Manuscripts26 Jan 202666.8$0.062
- T0232Blacklist Cards26 Jan 202693.0$0.13
- T0448Book Advert XML Repair25 Jan 202620.9$0.56
- T0342Company Lists25 Jan 202653.7$0.099
- T0341Company Lists25 Jan 202672.5$0.12
- T0160Library Cards24 Jan 202684.3$3.80
- T0139Bibliographic Data24 Jan 202675.2$0.079
- T0083Fraktur Adverts24 Jan 202671.7$0.075
- T0069Business Letters24 Jan 2026 · 32 signed letters59.0$0.19
- T0068Business Letters24 Jan 2026 · 25 unsigned letters43.0$0.11
- T0067Business Letters24 Jan 2026 · all 57 letters53.0$0.30
- T0448Book Advert XML Repair8 Dec 202517.9$0.30
- T0342Company Lists28 Oct 202546.5$0.088
- T0341Company Lists28 Oct 202563.4$0.11
- T0273Medieval Manuscripts24 Oct 202563.0$0.053
- T0232Blacklist Cards24 Oct 202592.2$0.11
- T0160Library Cards1 Oct 202585.0$2.70
- T0139Bibliographic Data1 Oct 202595.1$0.095
- T0160Library Cards30 Sep 202585.5$2.76
- T0083Fraktur Adverts30 Sep 202550.4$0.078
- T0069Business Letters30 Sep 2025 · 32 signed letters58.0$0.14
- T0068Business Letters30 Sep 2025 · 25 unsigned letters41.0$0.091
- T0067Business Letters30 Sep 2025 · all 57 letters51.0$0.23
- T0160Library Cards2 Sep 202589.4–
- T0139Bibliographic Data27 Aug 202591.3–
- T0083Fraktur Adverts9 May 202544.7–
- T0083Fraktur Adverts7 May 20259.1–
- T0069Business Letters17 Apr 2025 · 32 signed letters59.0–
- T0068Business Letters17 Apr 2025 · 25 unsigned letters41.0–
- T0067Business Letters17 Apr 2025 · all 57 letters52.0–