| Gemini 1.5: Unlocking multimodal understanding across millions of tokens of context | ✓ Link | 87.5 | Gemini 1.5 Pro (8-shot) | 2024-03-08 |
| PaLM 2 Technical Report | ✓ Link | 87.0 | PaLM 2 (few-shot, k=8, SC) | 2023-05-17 |
| Qwen3 Technical Report | ✓ Link | 83.53 | Qwen3-235B-A22B (Base) | 2025-05-14 |
| Encoder-Decoder Gemma: Improving the Quality-Efficiency Trade-Off via Adaptation | ✓ Link | 80.7 | T5Gemma 9B-9B PrefixLM IT | 2025-04-08 |
| DeepSeek-V4.1-Flash: Pushing the Limits of KV Cache Compression | ✓ Link | 80.2 | DeepSeek-V4.1-Flash Base | 2026-09-17 |
| Gemini: A Family of Highly Capable Multimodal Models | ✓ Link | 79.0 | Gemini Ultra (8-shot) | 2023-12-19 |
| Falcon-H1: A Family of Hybrid-Head Language Models Redefining Efficiency and Performance | ✓ Link | 76.33 | Falcon-H1-34B-Instruct (8-shot) | 2025-07-30 |
| PaLM 2 Technical Report | ✓ Link | 72.2 | PaLM 2 (8-shot, CoT) | 2023-05-17 |
| Scaling Instruction-Finetuned Language Models | ✓ Link | 72.0 | Flan-PaLM 540B (8-shot, fine-tuned, CoT + SC) | 2022-10-20 |
| Scaling Instruction-Finetuned Language Models | ✓ Link | 60.4 | Flan-U-PaLM 540B (CoT) | 2022-10-20 |
| Scaling Instruction-Finetuned Language Models | ✓ Link | 57.0 | Flan-PaLM 540B (8-shot, fine-tuned, CoT) | 2022-10-20 |
| PaLM: Scaling Language Modeling with Pathways | ✓ Link | 55.0 | PaLM 540B | 2022-04-05 |
| Transcending Scaling Laws with 0.1% Extra Compute | | 49.9 | U-PaLM 540B (CoT) | 2022-10-20 |
| Scaling Instruction-Finetuned Language Models | ✓ Link | 36 | text-davinci-003 | 2022-10-20 |
| Scaling Instruction-Finetuned Language Models | ✓ Link | 35 | code-davinci-002 | 2022-10-20 |
| Scaling Instruction-Finetuned Language Models | ✓ Link | 23.7 | text-davinci-002 | 2022-10-20 |
| Scaling Instruction-Finetuned Language Models | ✓ Link | 21.2 | Flan-PaLM 540B (8-shot, fine-tuned) | 2022-10-20 |
| Scaling Instruction-Finetuned Language Models | ✓ Link | 5.7 | GPT-3 Davinci 175B | 2022-10-20 |