ChartQA
Ergebnisse von KI-Benchmarks und Modellleistung
Typ: percentage
Erfasste Ergebnisse
41
Durchschnittspunktzahl
85.69%
Punktzahlbereich
76.10% - 90%
Führendes Modell
90% - Llama 4 Maverick
Punktzahlen im Zeitverlauf
Einzelne Benchmark-Punktzahlen nach Datum.
Modellergebnisse
| Organisation | Modelle | Gemeldet am | Höchste Punktzahl | Info | Selbst gemeldet | Quelle |
|---|---|---|---|---|---|---|
| Llama 4 Maverick | 05. Apr. 2025 | 90 % | - | Ja | Quelle | |
| Qwen 2.5 Math 72B | - | 89,50 % | inferred high-confidence family alias from qwen2.5-vl-72b (score=0.4700; benches=30) | Ja | Quelle | |
| Qwen 2.5 Math PRM 72B | - | 89,50 % | inferred family alias from qwen2.5-vl-72b (score=0.4050; benches=30) | Ja | Quelle | |
| Qwen 2.5 Math RM 72B | - | 89,50 % | inferred family alias from qwen2.5-vl-72b (score=0.4092; benches=30) | Ja | Quelle | |
| Qwen 2.5 VL 72B Instruct | 26. Jan. 2025 | 89,50 % | - | Ja | Quelle | |
| Llama 4 Scout | 05. Apr. 2025 | 88,80 % | - | Ja | Quelle | |
| Qwen 2 VL 2B | - | 88,30 % | inferred family alias from qwen2-vl-72b (score=0.4083; benches=15) | Ja | Quelle | |
| Qwen 2 VL 72B | - | 88,30 % | - | Ja | Quelle | |
| Pixtral Large | 18. Nov. 2024 | 88,10 % | - | Ja | Quelle | |
| Mistral Small 3.2 | 20. Juni 2025 | 87,40 % | - | Ja | Quelle | |
| Qwen 2.5 0.5B | - | 87,30 % | inferred family alias from qwen2.5-vl-7b (score=0.3131; benches=32) | Ja | Quelle | |
| Qwen 2.5 1.5B | - | 87,30 % | inferred family alias from qwen2.5-vl-7b (score=0.3131; benches=32) | Ja | Quelle | |
| Qwen 2.5 3B | - | 87,30 % | inferred family alias from qwen2.5-vl-7b (score=0.3531; benches=32) | Ja | Quelle | |
| Qwen 2.5 VL 7B Instruct | - | 87,30 % | - | Ja | Quelle | |
| DeepSeek VL2 | 13. Dez. 2024 | 86 % | - | Ja | Quelle | |
| GPT 4o (2024-08-06) | 06. Aug. 2024 | 85,70 % | - | Ja | Quelle | |
| GPT 4o Audio Preview (2025-06-03) | 03. Juni 2025 | 85,70 % | inferred modality/version alias from gpt-4o-2024-08-06 | Ja | Quelle | |
| GPT 4o Audio Preview (2024-10-01) | 01. Okt. 2024 | 85,70 % | inferred modality/version alias from gpt-4o-2024-08-06 | Ja | Quelle | |
| GPT 4o Audio Preview (2024-12-17) | 17. Dez. 2024 | 85,70 % | inferred modality/version alias from gpt-4o-2024-08-06 | Ja | Quelle | |
| GPT 4o Realtime Preview (2025-06-03) | 03. Juni 2025 | 85,70 % | inferred modality/version alias from gpt-4o-2024-08-06 | Ja | Quelle | |
| GPT 4o Realtime Preview (2024-10-01) | 01. Okt. 2024 | 85,70 % | inferred modality/version alias from gpt-4o-2024-08-06 | Ja | Quelle | |
| GPT 4o Search Preview | 11. März 2025 | 85,70 % | inferred modality/version alias from gpt-4o-2024-08-06 | Ja | Quelle | |
| GPT 4o Transcribe | 20. März 2025 | 85,70 % | inferred modality/version alias from gpt-4o-2024-08-06 | Ja | Quelle | |
| GPT 4o Transcribe Diarize | 15. Okt. 2025 | 85,70 % | inferred modality/version alias from gpt-4o-2024-08-06 | Ja | Quelle | |
| Llama 3.2 90B Vision Instruct | - | 85,50 % | inferred high-confidence family alias from llama-3.2-90b-instruct (score=0.5873; benches=13) | Ja | Quelle |
