AI Battle > Rankings > Humanity's Last Exam: Output Tokens per Task

Humanity's Last Exam: Output Tokens per Task Ranking — Frontier Evals (2026)

Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback) by Anthropic leads the Humanity's Last Exam: Output Tokens per Task ranking with a score of 451. The ranking covers 20 evaluated frontier evals from a public dataset that AI Battle re-checks daily.

Last updated: . Source: public benchmark datasets and provider pricing, aggregated daily; 20 entries shown.

Claude Fable 5.1 (Adaptive Reasoning, Medium451GPT-6 Astra (xhigh)280GPT-6 Astra (max)276Claude Fable 5.1 (Adaptive Reasoning, High E506Gemini 3.5 Flash-Lite367Muse Glimmer (high)642GPT-5.6 Sol (max)218GPT-5.6 Terra (max)216Claude Fable 5 (Adaptive Reasoning, Max Effo580Grok 4.6 (high)250Claude Opus 5 (Adaptive Reasoning, Xhigh Eff760Claude Fable 5.1 (Adaptive Reasoning, Xhigh 607MiniMax-M3466gpt-oss-120b (high)634K2 Horizon 375B A23B181GPT-5.6 Luna (max)237Gemini 3.8 Flash (high)606Kimi K3 (max)240Claude Opus 5 (Adaptive Reasoning, Max Effor841Mistral Medium 3.5331
Every entry in the dataset, scaled from the smallest to the largest published value.

All 20 entries in this dataset

#EntryCreatorValueΔ vs #1
1Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback)Anthropic4510.0%
2GPT-6 Astra (xhigh)OpenAI280-37.9%
3GPT-6 Astra (max)OpenAI276-38.8%
4Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback)Anthropic50612.2%
5Gemini 3.5 Flash-LiteGoogle367-18.6%
6Muse Glimmer (high)Meta64242.4%
7GPT-5.6 Sol (max)OpenAI218-51.7%
8GPT-5.6 Terra (max)OpenAI216-52.0%
9Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)Anthropic58028.6%
10Grok 4.6 (high)SpaceXAI250-44.6%
11Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)Anthropic76068.6%
12Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback)Anthropic60734.8%
13MiniMax-M3MiniMax4663.5%
14gpt-oss-120b (high)OpenAI63440.7%
15K2 Horizon 375B A23BMBZUAI Institute of Foundation Models181-60.0%
16GPT-5.6 Luna (max)OpenAI237-47.5%
17Gemini 3.8 Flash (high)Google60634.5%
18Kimi K3 (max)Kimi240-46.8%
19Claude Opus 5 (Adaptive Reasoning, Max Effort)Anthropic84186.7%
20Mistral Medium 3.5Mistral331-26.5%

How the values are distributed

Across the 20 numeric entries in this window the median is 409 and the mean is 434, so the typical entry sits 42.0 below the leader. The window spans 181 to 841.

63131402
8 equal buckets between the smallest (181) and largest (841) value; numbers under bars are entry counts.

10 of 20 entries are at or above the median, and the top three hold 26% of the combined value — read it against the metric below, since a higher number is not "better" for every dataset.

Creators in this ranking

9 creators appear in this window; the three most-represented hold 70% of the entries.

CreatorEntriesHighest valueTop entry
Anthropic6841Claude Opus 5 (Adaptive Reasoning, Max Effort)
OpenAI6634gpt-oss-120b (high)
Google2606Gemini 3.8 Flash (high)
Meta1642Muse Glimmer (high)
MiniMax1466MiniMax-M3
Mistral1331Mistral Medium 3.5
SpaceXAI1250Grok 4.6 (high)
Kimi1240Kimi K3 (max)
MBZUAI Institute of Foundation Models1181K2 Horizon 375B A23B

Where these models appear in other rankings

The same models are measured across multiple independent datasets. Aggregating the 20 entries above against every other AI Battle dataset gives this cross-section, computed by AI Battle and published nowhere upstream:

ModelDatasetsBest placements elsewhere
Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback)7#1/20 in AA-LCR v1.1: Output Tokens per Task, #2/20 in Humanity's Last Exam: Time per Task, #3/20 in AA-LCR v1.1: Score
GPT-6 Astra (xhigh)15#1/20 in Terminal-Bench v4.0: Score, #3/20 in AA-Omniscience Index: Score, #3/19 in MMMU-Pro: Token Usage
GPT-6 Astra (max)51#1/19 in MMMU-Pro: Score, #2/20 in AA-Omniscience Index, #2/20 in Artificial Analysis Intelligence Index by Open Weights / Proprietary
Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback)22#1/20 in SciCode: Output Tokens per Task, #2/20 in AA-Omniscience Index: Token Usage, #2/20 in AA-LCR v1.1: Output Tokens per Task
Gemini 3.5 Flash-Lite44#1/20 in End-to-End Response Time, #1/20 in Output Speed, #1/20 in Time per Intelligence Index Task
Muse Glimmer (high)43#1/20 in Output Tokens per Intelligence Index Task, #1/20 in Cost per Intelligence Index Task, #1/20 in Cost per Intelligence Index Task
GPT-5.6 Sol (max)56#3/20 in SciCode: Output Tokens per Task, #4/20 in 𝜏³-Banking: Output Tokens per Task, #5/20 in AA-LCR v1.1: Output Tokens per Task
GPT-5.6 Terra (max)52#7/20 in AA-Briefcase Output Tokens per Task, #9/20 in Output Speed, #9/20 in Output Speed
Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback)40#2/20 in SciCode: Score, #3/20 in AA-Omniscience Index, #3/20 in AA-Omniscience Index
Grok 4.6 (high)61#2/20 in 𝜏³-Banking: Score, #5/20 in AA-Briefcase Elo, #5/20 in AA-Omniscience Index
Claude Opus 5 (Adaptive Reasoning, Xhigh Effort)11#4/20 in AA-Briefcase Rubric Pass Rate by File Type (Normalized), #4/20 in AA-Briefcase Elo, #4/20 in GDPval-AA v2 Leaderboard
Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback)20#2/20 in Humanity's Last Exam: Score, #2/20 in AA-Omniscience Accuracy, #2/20 in AA-Briefcase Rubric Pass Rate by File Type (Normalized)
MiniMax-M357#1/20 in AA-Omniscience Hallucination Rate, #1/20 in 𝜏³-Banking: Output Tokens per Task, #4/20 in Pricing: Cache Hit, Input, and Output
gpt-oss-120b (high)39#1/20 in Cost to Run Artificial Analysis Intelligence Index, #1/20 in Cost to Run Artificial Analysis Intelligence Index, #1/20 in Humanity's Last Exam: Cost per Task
K2 Horizon 375B A23B24#2/20 in AA-Omniscience Hallucination Rate, #8/20 in AA-Briefcase Tool Calls Breakdown, Avg per Task, #12/20 in AA-Omniscience Index: Token Usage
GPT-5.6 Luna (max)58#1/11 in Cost per Task, #1/11 in Cost per Task, #3/20 in Pricing: Cache Hit, Input, and Output
Gemini 3.8 Flash (high)60#1/11 in Speed, #1/11 in Speed, #2/20 in Output Speed
Kimi K3 (max)60#1/20 in Context Window, #1/20 in AA-LCR v1.1: Score, #1/14 in Model Size: Total and Active Parameters
Claude Opus 5 (Adaptive Reasoning, Max Effort)44#2/20 in AA-Briefcase Elo, #2/20 in GDPval-AA v2 Leaderboard, #2/20 in AA-Briefcase Elo
Mistral Medium 3.541#2/20 in GDPval-AA v2: Output Tokens per Task, #3/20 in Output Tokens per Intelligence Index Task, #3/20 in Output Tokens per Intelligence Index Task

Metric reference

This ranking carries the metrics answer, reasoning from the Humanity's Last Exam: Output Tokens per Task dataset in the tab::eval-hle family. Units and the direction of "better" are not declared by the source dataset, so AI Battle does not invent them: values are shown exactly as published, and the ordering is the source dataset's own.

How this ranking is built

Rebuilt on from the public dataset Humanity's Last Exam: Output Tokens per Task (family tab::eval-hle). The source lists 20 entries and all 20 are shown here. AI Battle requests the dataset's rows, keeps the ordering the source publishes, and computes only the statistics on this page: median, mean, spread, the #1–#2 gap, the creator breakdown and the raw-data file. Nothing on this page is generated text about the entries — the numbers come from the source.

Refresh cadence: every deploy re-pulls the dataset, so this page tracks the source's own update cycle. Corrections: see the methodology page.

Limitations

Raw data

Machine-readable copy of this table, with the statistics below and the source provenance: /data/rankings/eval-hle-humanity-s-last-exam-output-tokens-per-task.json. Free to reuse with attribution to AI Battle and the source dataset.

FAQ — Humanity's Last Exam: Output Tokens per Task

What is the best model for Humanity's Last Exam: Output Tokens per Task in 2026?

Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback) by Anthropic currently leads the Humanity's Last Exam: Output Tokens per Task ranking with a score of 451, the highest value in this dataset across Frontier Evals. AI Battle aggregates the dataset from its public source and re-checks it daily.

How often is the Humanity's Last Exam: Output Tokens per Task ranking updated?

The Humanity's Last Exam: Output Tokens per Task ranking is rebuilt daily from the upstream public dataset. Scores are the source dataset's own published values, so positions track whatever that source last released — each ranking page names the dataset it comes from.

Which models rank in the top 3 for Humanity's Last Exam: Output Tokens per Task?

The top 3 are: 1. Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback) (451), 2. GPT-6 Astra (xhigh) (280), 3. GPT-6 Astra (max) (276).

All AI Battle Rankings

Directories