Benchmark Almanac

EGI-BENCH

GPT-5.5 65.6 leads since October 2026
Not in model cards

Agents work across eight enterprise evaluations. The suite combines their results into six scores covering activities such as planning, retrieving evidence, and using tools.

Released
October 2026
Kind
Simulated companies
Level
University
Scored
Checked automatically
Answers
Agent task
Tasks
All public
Measures
Capability
Language
English
In and out
Text, Documents, Code → Text, Actions, Code
Interaction
Agentic
Environment
APIs and tools
Refresh
One fixed set
Run by
Anyone
Strongest model tested
GPT-5.5 #21 on the capability index
Last model added
April 2026
Scores refreshed
November 20th, 2026
14
Entries in its paper, 12 models we track

GPT-5.5 leads with 65.6 among models released by April 2026 GPT-5.5 leads with 65.6 among models released by April 2026

Best score per entry from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. 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. This view shows only models we track, each at its best.

The newest model with a score here came out in April 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 GPT-5.5 quick 65.6
  2. 2 Claude Sonnet 4.5 quick 65.3
  3. 3 Gemini 3.1 Pro quick 63.5
  4. 4 Claude Opus 4.5 quick 63.3
  5. 5 GPT-4o (snapshot unspecified) quick 59.4
  6. 6 Claude Haiku 4.5 quick 59.1
  7. 7 Gemini 2.5 Pro (snapshot unspecified) quick 59.1
  8. 8 GPT-4.1 quick 58.3
  9. 9 GPT-5 mini quick 51.9
  10. 10 Gemini 2.5 Flash Sep 2025, quick 51.4
  11. 11 Claude 3.5 Haiku quick 50.8
  12. 12 Gemini 2.0 Flash Feb 2025, quick 49.3
  13. 13 GPT-5 nano quick 45.1
  14. 14 GPT-4o mini quick 39.6

Source: The EGI-BENCH paper

#EntryScoreSettingDateSource
1 GPT-5.5 65.6 quick Apr 2026 Official leaderboard
2 Claude Sonnet 4.5 65.3 quick Sep 2025 Official leaderboard
3 Gemini 3.1 Pro 63.5 quick Feb 2026 Official leaderboard
4 Claude Opus 4.5 63.3 quick Nov 2025 Official leaderboard
5 GPT-4o (snapshot unspecified) 59.4 quick Official leaderboard
6 Claude Haiku 4.5 59.1 quick Oct 2025 Official leaderboard
7 Gemini 2.5 Pro (snapshot unspecified) 59.1 quick Official leaderboard
8 GPT-4.1 58.3 quick Apr 2025 Official leaderboard
9 GPT-5 mini 51.9 quick Aug 2025 Official leaderboard
10 Gemini 2.5 Flash Sep 2025 51.4 quick Sep 2025 Official leaderboard
11 Claude 3.5 Haiku 50.8 quick Oct 2024 Official leaderboard
12 Gemini 2.0 Flash Feb 2025 49.3 quick Feb 2025 Official leaderboard
13 GPT-5 nano 45.1 quick Aug 2025 Official leaderboard
14 GPT-4o mini 39.6 quick Jul 2024 Official leaderboard
#ModelScoreSettingDateSource
1 GPT-5.5 65.6 quick Apr 2026 Official leaderboard
2 Claude Sonnet 4.5 65.3 quick Sep 2025 Official leaderboard
3 Gemini 3.1 Pro 63.5 quick Feb 2026 Official leaderboard
4 Claude Opus 4.5 63.3 quick Nov 2025 Official leaderboard
5 Claude Haiku 4.5 59.1 quick Oct 2025 Official leaderboard
6 GPT-4.1 58.3 quick Apr 2025 Official leaderboard
7 GPT-5 mini 51.9 quick Aug 2025 Official leaderboard
8 Gemini 2.5 Flash Sep 2025 51.4 quick Sep 2025 Official leaderboard
9 Claude 3.5 Haiku 50.8 quick Oct 2024 Official leaderboard
10 Gemini 2.0 Flash Feb 2025 49.3 quick Feb 2025 Official leaderboard
11 GPT-5 nano 45.1 quick Aug 2025 Official leaderboard
12 GPT-4o mini 39.6 quick Jul 2024 Official leaderboard

Rankings by part

We read 6 parts from the benchmark’s paper, each ranking models on one part of the test, such as M1: Task Outcome Accuracy, M2: Tool Actioning Reliability and M3: Planning & Workflow. GPT-5.5 leads 3 of them, and 2 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

M1: Task Outcome Accuracy

12 models we track have a score here.

  1. 1 GPT-5.5 83.0
  2. 2 Gemini 3.1 Pro ↑1 78.5
  3. 3 Claude Opus 4.5 ↑1 74.3
  4. 4 Claude Sonnet 4.5 ↓2 69.9
  5. 5 GPT-4.1 ↑1 69.9
  6. 6 Claude Haiku 4.5 ↓1 69.0
  7. 7 Gemini 2.5 Flash Sep 2025 ↑1 59.3
  8. 8 GPT-5 mini ↓1 59.1
  9. 9 Gemini 2.0 Flash Feb 2025 ↑1 48.6
  10. 10 Claude 3.5 Haiku ↓1 48.2

Number one over time

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

OpenAI
Benchmark paper Not tested since Oct 2026
GPT-5.5 65.6

Source: The EGI-BENCH 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
30.0 40.0 50.0 60.0 70.0 2025 2026 30.0 40.0 50.0 60.0 70.0 2025 2026

Source: Official EGI-BENCH 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 Economic work

All 210 in Economic work
  • JobBench In use 130 tasks from 35 occupations workers most want delegated. Agent works through messy reference files, graded by fact-anchored binary rubrics. GLM-5.3 61.4%
  • GDPval In use Real work tasks from 44 occupations with reference files. Experts blindly compare model deliverables with expert work. Metric is win-or-tie rate. Seed2.1 Pro 87.9%
  • GDPval-AA v2.1 In use Agents with shell and web access complete 220 GDPval gold tasks from 44 occupations. Blind pairwise comparisons give Elo ratings. Claude Opus 5.5 1,846
  • AA-Briefcase v1.1 In use Version 1.1 of Artificial Analysis's multi-hour office work evaluation, with Elo ratings fitted by Crowd-BT. Claude Opus 5.5 1,822
  • Agents' Last Exam v1.1 In use Agents complete 122 public long-horizon professional workflow tasks across 55 sub-industries. Deterministic graders check outcomes against hidden references. GPT-6.2 Astra 47.5%
  • CoWorkBench In use Agent completes long-horizon productivity tasks across computer science, finance, law and medicine. Qwen in-house scoring (method unpublished). Claude Fable 5 75.9%

Data audit

8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 7 months ago, in April 2026, for GPT-5.5. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
  • Pass Independent runs Scores from the benchmark’s paper (14 entries).
  • Pass Models scored 12 models we track, plus 2 other entries on its leaderboard.
  • Warning Newest score The newest came 7 months ago, in April 2026, for GPT-5.5.
  • Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5.5 (#21).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on arxiv.org: 12 rows for models we track and 2 other entries at the last read.
  • Pass Last read November 20th, 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