Benchmark Almanac

SpireClimb

GPT-6 Astra 25.7% leads since Oct 2026 Not in model cards

Agents play Slay the Spire 1 and 2 at max difficulty through a computer-use harness on fixed seeds. Score is the mean win rate.

Top human players win ~85% (StS1) and ~95% (StS2). Seeds are not published.

Made by Epoch AI Released Oct 2026 Video games Tasks kept private Last model added Sep 2026 Scores refreshed 31 Oct 2026
Level
Expert Specialists with years of training, often over hours of work
Scored
Checked automatically A program checks the result against an answer key, unit tests or the expected final state.
Answers
Agent task Many steps in a terminal, browser, desktop, app, game or robot.
Paper Leaderboard
7
Models scored in Epoch AI data

GPT-6 Astra leads with 25.7%

Best score per model in Epoch AI’s data. Where a model ran with several settings or harnesses, we keep its best.

  1. 25.7 GPT-6 Astra
  2. 24.7 GPT-6.1 Sol
  3. 7.50 GPT-6 Sol
  4. 3.33 Claude Opus 5.5
  5. 2.00 Claude Fable 5.1
  6. 1.33 Claude Sonnet 5.5
  7. 0.0 GPT-6 Luna
  1. 1 GPT-6 Astra max 25.7%
  2. 2 GPT-6.1 Sol max 24.7%
  3. 3 GPT-6 Sol max 7.50%
  4. 4 Claude Opus 5.5 max 3.33%
  5. 5 Claude Fable 5.1 max 2.00%
  6. 6 Claude Sonnet 5.5 max 1.33%
  7. 7 GPT-6 Luna max 0%
#ModelScoreSettingDateSource
1 GPT-6 Astra 25.7% max Sep 2026 Epoch AI
2 GPT-6.1 Sol 24.7% max Sep 2026 Epoch AI
3 GPT-6 Sol 7.50% max Sep 2026 Epoch AI
4 Claude Opus 5.5 3.33% max Sep 2026 Epoch AI
5 Claude Fable 5.1 2.00% max Sep 2026 Epoch AI
6 Claude Sonnet 5.5 1.33% max Sep 2026 Epoch AI
7 GPT-6 Luna 0% max Sep 2026 Epoch AI

Number one over time

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

OpenAI
Epoch AI runs 1 leader
GPT-6 Astra 25.7%

Scores over time

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

Run or collected by Epoch AI Best so far
0% 20% 40% 60% 80% 100% Near the ceiling (85%) GPT-6 Astra Claude Fable 5.1 0% 20% 40% 60% 80% 100% Near the ceiling (85%) GPT-6 Astra Claude Fable 5.1

No lab model card reports it yet

We read the headline results of 479 launch posts from 51 labs, and none of them list this benchmark.

More in Games and long tasks

All 68 in Games and long tasks
  • ARC-AGI-3 In use Interactive game environments where agents must explore, infer goals and rules, and act efficiently without instructions. GPT-6 Astra 100%
  • MineDojo Fading Agents complete open-ended tasks in Minecraft, such as mining blocks and fighting mobs, from a large simulation suite. Programmatic checks score many tasks. Seed2.0 Pro 49%
  • Orak Fading LLM agents play 12 popular video games across most genres through an MCP interface. Scores come from in-game progress or outcomes per title. Nemotron 3 Nano 4B 22.9
  • GameWorld Only in system cards Multimodal game agents play 34 browser games across 170 tasks. Outcomes are checked from the game state. Gemini 3 Flash 41.9%
  • Swarm Bench Only in system cards Models orchestrate agent swarms on complex tasks that benefit from coordinated decomposition and parallel execution. Kimi K3 76.3%
  • TextQuests Only in system cards Agents play 25 classic interactive fiction games. Measures game progress and harmful actions per 100 steps. Muse Spark 2.40