Benchmark Almanac

ALCE

ChatGPT 84.8% tops its leaderboard Stable Beluga 2 63.6% leads the models we track
Not in model cards

Models answer questions and cite passages. ALCE evaluates answer quality and checks whether the citations support the answer.

The Princeton repository has 532 GitHub stars. Appendix F describes the human evaluation that Surge AI ran.

Released
May 2023
Kind
Questions that need search
Level
Everyday
Scored
Graded by a model
Answers
Open-ended
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Single turn
Items from
Another benchmark
Refresh
One fixed set
Run by
Anyone
Strongest model tested
Stable Beluga 2 #205 on the capability index
Last model added
July 2023
Scores refreshed
November 21st, 2026
14
Entries in its paper, 4 models we track

ChatGPT tops the leaderboard with 84.8% Stable Beluga 2 leads with 63.6% among models released by July 2023

Best score per entry from the benchmark's paper, read on November 21st, 2026. A paper’s table doesn’t add newer models. The paper reports several metrics, so this leaderboard uses the ASQA citation-recall column as its headline. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders. The best model we track is Stable Beluga 2, at 63.6%. This view shows only models we track, each at its best.

The newest model with a score here came out in July 2023, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 ChatGPT w/ Rerank 84.8%
  2. 2 ChatGPT-16K Vanilla (5-psg) 76.2%
  3. 3 GPT-4 (snapshot unspecified) Vanilla (20-psg) 73%
  4. 4 Vicuna-13B w/ Rerank 71.9%
  5. 5 Oasst-33B w/ Rerank 66.7%
  6. 6 Stable Beluga 2 Vanilla (5-psg) 63.6%
  7. 7 LLaMA-2-70B-Chat Vanilla (5-psg) 62.9%
  8. 8 LLaMA-2-7B-Chat Vanilla (5-psg) 50.9%
  9. 9 LLaMA-33B w/ Rerank 41.3%
  10. 10 Vicuna-7B Vanilla (3-psg) 40.3%
  11. 11 LLaMA-2-13B-Chat Vanilla (5-psg) 38.4%
  12. 12 LLaMA-13B w/ Rerank 28.1%
  13. 13 Alpaca-7B Vanilla (3-psg) 12.3%
  14. 14 LLaMA-7B Vanilla (3-psg) 6.20%

Source: The ALCE paper

#EntryScoreSettingDateSource
1 ChatGPT 84.8% w/ Rerank Official leaderboard
2 ChatGPT-16K 76.2% Vanilla (5-psg) Official leaderboard
3 GPT-4 (snapshot unspecified) 73% Vanilla (20-psg) Official leaderboard
4 Vicuna-13B 71.9% w/ Rerank Official leaderboard
5 Oasst-33B 66.7% w/ Rerank Official leaderboard
6 Stable Beluga 2 63.6% Vanilla (5-psg) Jul 2023 Official leaderboard
7 LLaMA-2-70B-Chat 62.9% Vanilla (5-psg) Official leaderboard
8 LLaMA-2-7B-Chat 50.9% Vanilla (5-psg) Official leaderboard
9 LLaMA-33B 41.3% w/ Rerank Feb 2023 Official leaderboard
10 Vicuna-7B 40.3% Vanilla (3-psg) Official leaderboard
11 LLaMA-2-13B-Chat 38.4% Vanilla (5-psg) Official leaderboard
12 LLaMA-13B 28.1% w/ Rerank Feb 2023 Official leaderboard
13 Alpaca-7B 12.3% Vanilla (3-psg) Official leaderboard
14 LLaMA-7B 6.20% Vanilla (3-psg) Feb 2023 Official leaderboard
#ModelScoreSettingDateSource
1 Stable Beluga 2 63.6% Vanilla (5-psg) Jul 2023 Official leaderboard
2 LLaMA-33B 41.3% w/ Rerank Feb 2023 Official leaderboard
3 LLaMA-13B 28.1% w/ Rerank Feb 2023 Official leaderboard
4 LLaMA-7B 6.20% Vanilla (3-psg) Feb 2023 Official leaderboard

2 models have held number one in the benchmark’s paper since its release in May 2023

Lab claims count from the day the card came out, independent runs from the model’s release.

Meta Stability AI
Benchmark paper Not tested since Jul 2023

Source: The ALCE paper

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
In the benchmark’s paper Best so far
0% 20% 40% 60% 80% 100% 2023 Near the ceiling (85%) Stable Beluga 2 LLaMA-33B 0% 20% 40% 60% 80% 100% 2023 Near the ceiling (85%) Stable Beluga 2 LLaMA-33B

Source: Official ALCE leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Knowledge

All 332 in Knowledge
  • MMLU-Pro In use Harder 10-option version of MMLU with reasoning-focused questions across 14 subjects. Intern-S2-397B 89.8%
  • MMLU In use 57-subject multiple-choice exam questions from elementary to professional level. Scored by accuracy. Seed1.8 92.3%
  • SuperGPQA In use 26,529 graduate-level multiple-choice questions across 285 disciplines. Accuracy. Qwen 3.6 Max (Preview) 73.9%
  • WideSearch In use 200 bilingual broad info-seeking tasks: agent collects many atomic facts into a table. Scored by success rate and row/item F1. Tencent Hy4 preview 83.9%
  • BBEH micro average In use Models solve harder replacements for each of the 23 BIG-Bench Hard tasks, 4,520 examples in total. The score is the share of all examples answered correctly. Gemma 4 31B IT 74.4%
  • HLE-Diamond In use Models answer 1,000 cleaned expert questions from Humanity's Last Exam, 500 on reasoning and 500 on knowledge, closed book. Graded on exact answers. Claude Fable 5.2 77.9%

Data audit

9 of 10 checks pass, 1 warning. Look first at the Newest score check. The newest came 3 years ago, in July 2023, for Stable Beluga 2. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 3 pass 1 note
  • Pass Independent runs Scores from the benchmark’s paper (14 entries).
  • Pass Models scored 4 models we track, plus 10 other entries on its leaderboard.
  • Warning Newest score The newest came 3 years ago, in July 2023, for Stable Beluga 2.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Stable Beluga 2 (#205).
  • Pass Headroom The best is 63.6%, under 85%, so it still separates models.
Leaderboard import 3 pass
  • Pass Importer We read the board on arxiv.org: 12 rows for models we track and 10 other entries at the last read.
  • Pass Last read November 21st, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check