Agentic & Tool UseStill discriminating

Earthborne Rangers (EBR-bench)

Completing objectives in a complex board game after repeated practice and note-taking.

  • 31 models have a score, from Epoch AI.
  • The leader is claude-fable-5-2_max at 90.5, dated Nov 6, 2026.
  • The top 10 are within 45 points / 100 of each other.
  • Its newest result is dated Nov 6, 2026, and 4 results were published in the last 30 days.
  • A score on this page last changed on Nov 11, 2026.

Metric

index (points / 100)

Status

Still discriminating

Category

Agentic & Tool Use

Maintainer

Epoch AI

What it tests

Planning and sustained execution in an unfamiliar campaign game. Agents manage strategy, tactics, and written notes while pursuing objectives across the first five in-game days.

How it works

Default runs contain eight practice playthroughs followed by two scored playthroughs. The topline uses the better of the final two, measured as the share of 21 available objectives. Epoch also compares context-compaction settings for its canonical leaderboard.

How to read a score

Higher means more objectives achieved after practice. This is the topline performance score, not a measured learning gain; showing improvement from practice requires the source's separate controlled comparison.

Limitations

Best-of-two selection, note-taking, compaction, and the game harness influence the result. The benchmark covers a limited campaign segment and does not establish general on-the-job learning ability.

Leaderboard

Checked Nov 17, 2026, 07:00 UTC · scores last changed Nov 11, 2026

Best Earthborne Rangers (EBR-bench) score per model, ranked. Each score links to its source.

#ModelScoreSource
1Anthropic: Claude Fable 5.2(max)Anthropic90.5epoch
2Google: Gemini 4 Carbon(high)Google DeepMind87.6epoch
3OpenAI: GPT-6.1 Terra(max)OpenAI78.1epoch
4OpenAI: GPT-6 Astra(max)OpenAI76.2epoch
5Anthropic: Claude Opus 5.5(max)Anthropic71.4epoch
6Google: Gemini 4 Argon(high)Google DeepMind64.8epoch
7Anthropic: Claude Fable 5.1(max)Anthropic57.1epoch
8OpenAI: GPT-6.1 Sol(max)OpenAI54.3epoch
9OpenAI: GPT-6 Sol(max)OpenAI53.3epoch
10Anthropic: Claude Opus 5(max)Anthropic45.7epoch
11OpenAI: GPT-5.6 Sol(max)OpenAI44.8epoch
12Anthropic: Claude Haiku 5.5(max)Anthropic43.8epoch
13Anthropic: Claude Fable 5(max)Anthropic39.5epoch
14OpenAI: GPT-5.5(xhigh)OpenAI34.3epoch
15muse-wonder-1.0_max(max)Meta AI31.4epoch
16SpaceXAI: Grok 4.6(xhigh)xAI30.5epoch
17Anthropic: Claude Opus 4.8(max)Anthropic28.6epoch
18OpenAI: GPT-5.4(xhigh)OpenAI25.4epoch
19OpenAI: GPT-5.2(xhigh)OpenAI23epoch
20Anthropic: Claude Opus 4.7(max)Anthropic19epoch
21mistral-large-4-0-oss_high(high)Mistral AI18.1epoch
22claude-opus-4-5-20251101_128K(budget:128k)Anthropic14.3epoch
23Google: Gemini 3.1 Pro PreviewGoogle DeepMind14.3epoch
24Anthropic: Claude Opus 4.6(max)Anthropic12.7epoch
25OpenAI: GPT-5(high)OpenAI12.7epoch
26Z.ai: GLM 5.2(max)Z.ai (Zhipu AI)9.5epoch
27Qwen: Qwen3.7 MaxAlibaba9.5epoch
28Anthropic: Claude Opus 4.1Anthropic7.9epoch
29Google: Gemini 3.5 Flash(high)Google DeepMind4.8epoch
30MoonshotAI: Kimi K2.6Moonshot2.4epoch
31Anthropic: Claude Sonnet 4.5Anthropic2.4epoch

Run your own prompt on the top 3

Claude Fable 5.2, Gemini 4 Carbon, GPT-6.1 Terra, side by side on the same prompt, with each one's time and cost. A free account includes starter credit.

Compare them

Compare the leaders

More Agentic & Tool Use benchmarks

Earthborne Rangers (EBR-bench) is maintained by Epoch AI. Scores on this page are collected from openly-licensed sources and each links to its origin. See the full matrix to compare every model across every benchmark.