HLE Text-Only (Scale) leaderboard
As of 22 Nov 2026, the top score on HLE Text-Only (Scale) is 54.2% by GPT-6 Astra (OpenAI), ahead of Gemini 3.1 Pro at 47.3%, across 64 model configurations with a published result (Scale AI SEAL, result dated 9 Sep 2026). Humanity's Last Exam without its image questions, run by Scale AI: the only HLE score models without image input get. Not comparable with the full exam, which the BenchLeader Index uses. HLE Text-Only (Scale) is shown for reference only and does not count toward the BenchLeader Index.
Key facts (as of 22 Nov 2026)
- GPT-6 Astra holds the top HLE Text-Only (Scale) score at 54.2% (Scale AI SEAL, 9 Sep 2026).
- Gemini 3.1 Pro is second on HLE Text-Only (Scale) at 47.3% (Scale AI SEAL, 10 Apr 2026).
- HLE Text-Only (Scale): Humanity's Last Exam without its image questions, run by Scale AI: the only HLE score models without image input get. Not comparable with the full exam, which the BenchLeader Index uses (Scale AI SEAL).
- 64 model configurations have a published HLE Text-Only (Scale) result on BenchLeader (BenchLeader, 22 Nov 2026).
- HLE Text-Only (Scale) does not feed the BenchLeader Index; it is shown for reference (BenchLeader methodology).
Humanity's Last Exam without its image questions, run by Scale AI: the only HLE score models without image input get. Not comparable with the full exam, which the BenchLeader Index uses.
Scale AI SEAL marks 56 of these 64 results as possibly contaminated: the model was evaluated after the questions were public, so they may already be in its training data. Those scores are in the table below; this board does not feed the index.
- Published by
- Scale AI SEAL
- Category
- Reasoning
- Index weight
- Reference only
- Models
- 64
- Data as of
- 22 Nov 2026
- Newest result
- 12 Oct 2026
Scale AI SEAL Leaderboards.
- 1GPT-6 Astra54.2%
- 2Gemini 3.1 Pro (high)47.3%
- 3Claude Fable 5.1 (xhigh)46.8%
- 4Claude Sonnet 5.5 (xhigh)46.6%
- 5GPT-5.4 Pro45.3%
- 6Gemini 3.8 Flash45.0%
- 7Muse Spark40.9%
- 8Gemini 3 Pro37.7%
- 9GPT-5.4 (thinking)36.5%
- 10Claude Opus 4.6 (max)36.2%
- 11GPT-5 Pro33.3%
- 12Claude Haiku 5.530.4%
- 13GPT-5.228.5%
- 14Claude Opus 4.5 (thinking)26.3%
- 15GPT-526.3%
| # | |||||
|---|---|---|---|---|---|
| 1 | 54.2% | 68.0 | 2026-09-09 | possible | |
| 2 | 47.3% | 58.2 | 2026-04-10 | possible | |
| 3 | 46.8% | 63.9 | 2026-09-03 | possible | |
| 4 | 46.6% | 62.4 | 2026-10-12 | – | |
| 5 | 45.3% | 62.7 | 2026-03-23 | possible | |
| 6 | 45.0% | 55.7 | 2026-09-09 | possible | |
| 7 | 40.9% | 64.8 | 2026-04-08 | possible | |
| 8 | 37.7% | 60.3 | 2025-11-19 | possible | |
| 9 | 36.5% | – | 2026-03-10 | possible | |
| 10 | 36.2% | 56.8 | 2026-02-17 | possible | |
| 11 | 33.3% | 61.3 | 2025-11-06 | possible | |
| 12 | 30.4% | – | 2026-10-08 | – | |
| 13 | 28.5% | 55.4 | 2025-12-15 | possible | |
| 14 | 26.3% | 59.6 | 2025-11-26 | possible | |
| 15 | 26.3% | 56.8 | 2025-08-07 | possible | |
| 16 | 24.6% | – | 2025-11-26 | possible | |
| 17 | 22.1% | 53.5 | 2025-06-05 | possible | |
| 18 | 20.6% | 54.8 | 2025-04-16 | possible | |
| 19 | 19.8% | 53.1 | 2025-04-16 | possible | |
| 20 | 19.7% | 56.5 | 2025-08-22 | possible | |
| 21 | 19.4% | 57.1 | 2026-02-17 | possible | |
| 22 | 18.9% | 50.8 | 2025-04-16 | possible | |
| 23 | 18.4% | 59.7 | 2025-04-10 | possible | |
| 24 | 18.4% | 56.9 | 2025-05-06 | possible | |
| 25 | 15.5% | 46.2 | 2025-09-18 | possible | |
| 26 | 15.4% | 51.7 | 2025-08-05 | possible | |
| 27 | 14.5% | 51.4 | 2025-04-16 | possible | |
| 28 | 14.1% | 52.9 | 2025-10-02 | possible | |
| 29 | 14.0% | 49.0 | 2025-06-02 | possible | |
| 30 | 13.9% | 56.9 | 2025-11-26 | possible | |
| 31 | 13.4% | 48.3 | 2025-04-10 | possible | |
| 32 | 12.9% | 48.5 | 2025-09-08 | possible | |
| 33 | 12.6% | 52.3 | 2025-04-17 | possible | |
| 34 | 11.8% | 50.0 | 2025-05-01 | possible | |
| 35 | 11.3% | 58.0 | 2025-08-08 | possible | |
| 36 | 10.8% | 55.7 | 2025-05-24 | possible | |
| 37 | 10.3% | 46.2 | 2025-04-10 | possible | |
| 38 | 9.7% | 44.5 | 2025-09-18 | possible | |
| 39 | 9.6% | 51.6 | 2025-08-13 | possible | |
| 40 | 9.4% | 46.7 | 2025-08-13 | possible | |
| 41 | 8.5% | 47.0 | 2025-04-10 | – | |
| 42 | 8.0% | 52.5 | 2026-03-23 | possible | |
| 43 | 7.9% | 52.5 | 2025-04-10 | possible | |
| 44 | 7.8% | 53.6 | 2025-04-10 | – | |
| 45 | 7.7% | – | 2025-04-10 | possible | |
| 46 | 7.7% | 53.6 | 2025-10-02 | possible | |
| 47 | 7.6% | 53.8 | 2025-05-23 | possible | |
| 48 | 7.4% | 51.7 | 2025-08-08 | possible | |
| 49 | 6.5% | 43.4 | 2025-04-10 | – | |
| 50 | 6.5% | – | 2025-11-26 | possible | |
| 51 | 6.3% | 51.7 | 2025-05-23 | possible | |
| 52 | 5.8% | 50.3 | 2025-04-10 | possible | |
| 53 | 5.4% | 49.9 | 2025-05-23 | possible | |
| 54 | 5.3% | 41.7 | 2025-04-10 | possible | |
| 55 | 5.0% | 48.5 | 2025-04-14 | possible | |
| 56 | 4.7% | 51.0 | 2025-07-23 | possible | |
| 57 | 4.5% | 44.2 | 2025-04-10 | possible | |
| 58 | 4.5% | 43.7 | 2025-04-10 | – | |
| 59 | 4.4% | 39.1 | 2025-04-10 | possible | |
| 60 | 4.4% | 44.2 | 2025-05-13 | possible |
AI reasoning and math benchmark leaderboard · All LLM benchmarks
Related benchmarks
Top head-to-heads on HLE Text-Only (Scale)
Cite as: BenchLeader, “HLE Text-Only (Scale) leaderboard”, data as of 22 Nov 2026, https://www.benchleader.com/benchmarks/scale_hle_text_only.
HLE Text-Only (Scale): questions
- What does HLE Text-Only (Scale) measure?
- Humanity's Last Exam without its image questions, run by Scale AI: the only HLE score models without image input get. Not comparable with the full exam, which the BenchLeader Index uses. Scores are reported in percent of tasks solved; higher is better.
- What is the top score on HLE Text-Only (Scale)?
- As of 22 Nov 2026, the highest score on HLE Text-Only (Scale) is 54.2% by GPT-6 Astra (result dated 9 Sep 2026), ahead of Gemini 3.1 Pro at 47.3%. Scores come from Scale AI SEAL and are re-read daily.
- Which AI model leads HLE Text-Only (Scale)?
- GPT-6 Astra leads HLE Text-Only (Scale) with 54.2% as of 22 Nov 2026 (result dated 9 Sep 2026), ahead of Gemini 3.1 Pro at 47.3%.
- How many models have HLE Text-Only (Scale) results?
- 64 model configurations have a HLE Text-Only (Scale) result on BenchLeader, all taken from Scale AI SEAL.
- Who runs HLE Text-Only (Scale) and how often is it updated?
- HLE Text-Only (Scale) is published by Scale AI SEAL. BenchLeader re-reads the published results once a day and records the date each result was published.
- Does HLE Text-Only (Scale) count toward the BenchLeader Index?
- No. HLE Text-Only (Scale) is shown for reference but left out of the composite index.