Benchmark Almanac

SWE-Game

Claude Opus 5 64.0 leads since September 2026
Not in model cards

Agents build, complete, repair, or port games across 247 tasks derived from 41 reference games. Runtime probes and input replays check gameplay mechanics and preserved behavior. Vision-language rubrics separately grade presentation in construction tasks.

The repair track contains 83 fault cases. Other tracks contain 41 tasks each. Porting uses a separate Unity execution contract. A public task download was not verified.

Released
September 2026
Kind
Building whole projects
Level
University
Scored
Checked automatically
Answers
Code
Measures
Capability
Language
English
In and out
Text, Code, Documents → Code, Actions
Interaction
Agentic
Environment
Terminal, Godot, Unity
Items from
Real work and use
Refresh
One fixed set
Run by
A host
In views
Agents
Strongest model tested
Claude Opus 5 #11 on the capability index
Last model added
August 2026
Scores refreshed
November 20th, 2026

In the authors’ words

We introduce SWE-Game, a benchmark of 247 tasks grounded in 41 executable reference Godot games spanning 13 gameplay categories in 2D and 3D. Five task types cover development from a brief, implementation from a game design document, skeleton completion, repair of 83 injected-fault cases, and Godot-to-Unity porting. Reference materials specify the intended gameplay, while a shared instrumentation interface lets evaluator-owned drivers and probes execute actions and observe independently implemented games. Evaluation combines engine-state checks, certified reference-input replay, and agent-authored feature demonstrations to assess mechanic correctness, demonstrated playability, and behavioral restoration and preservation after repairs. Game-specific vision-language rubrics separately assess presentation. Across six models, Opus5 achieves the highest overall score in all five task types. Best overall scores remain below 60 out of 100 across the three construction tasks, with Brief-to-Game reaching 50.38. Analysis of reviewed submissions identifies requirement omissions and gameplay logic errors as predominant implementation problems. On human-labeled behaviors from 100 agent-built games, executable checks achieve 92.59% balanced accuracy, compared with 78.41% for a video-based VLM judge. Rubric-based visual scores reach a Spearman correlation of 0.829 with human ratings of 200 gameplay clips. Together, these results characterize current agent capabilities across game-development activities and support combining runtime evidence with visual assessment.

Abstract of SWE-Game: Can Coding Agents Build the Games We Want? (arXiv:2609.33678, CC0)
6
Models scored in its paper

Claude Opus 5 leads with 64.0 among models released by August 2026

Best score per model from the benchmark's paper, read on November 20th, 2026. A paper’s table doesn’t add newer models. The authors report no overall score, so this is the plain mean of the five track totals. Where a model ran with several settings or harnesses, we keep its best.

The newest model with a score here came out in August 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 Claude Opus 5 64.0
  2. 2 GPT-5.6 Luna 49.3
  3. 3 Grok 4.6 47.8
  4. 4 MiniMax-M3 37.0
  5. 5 Qwen3.8 Flash 36.0
  6. 6 GLM-5.3-Flash 35.4

Source: The SWE-Game paper

#ModelScoreSettingDateSource
1 Claude Opus 5 64.0 Jul 2026 Official leaderboard
2 GPT-5.6 Luna 49.3 Jul 2026 Official leaderboard
3 Grok 4.6 47.8 Aug 2026 Official leaderboard
4 MiniMax-M3 37.0 Jun 2026 Official leaderboard
5 Qwen3.8 Flash 36.0 Aug 2026 Official leaderboard
6 GLM-5.3-Flash 35.4 Aug 2026 Official leaderboard

Rankings by part

We read 5 parts from the benchmark’s paper, each ranking models on one part of the test, such as Brief-to-Game, GDD-to-Game and Skeleton Completion. Claude Opus 5 leads every one of them. Arrows show how far a model moved from its overall rank among the models we track.

Brief-to-Game

6 models we track have a score here.

  1. 1 Claude Opus 5 50.2
  2. 2 Grok 4.6 ↑1 38.2
  3. 3 GPT-5.6 Luna ↓1 32.6
  4. 4 GLM-5.3-Flash ↑2 30.3
  5. 5 MiniMax-M3 ↓1 29.5
  6. 6 Qwen3.8 Flash ↓1 26.0

Number one over time

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

Anthropic
Benchmark paper Not tested since Sep 2026

Source: The SWE-Game 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 2026 30.0 40.0 50.0 60.0 70.0 2026

Source: Official SWE-Game 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 Coding

All 523 in Coding
  • SWE-bench Verified In use 500 human-validated GitHub issues from 12 Python repos. Agent writes a patch that must pass hidden tests. Claude Opus 4.7 83.5%
  • LiveCodeBench v6 In use Competitive programming problems collected after model cutoffs from LeetCode, AtCoder, Codeforces. Scored pass@1 on hidden tests. DeepSeek-V4-Pro 93.5%
  • SWE-bench Multilingual In use 300 GitHub-issue tasks across nine programming languages in SWE-bench format. Patch graded by repository tests. Claude Fable 5.2 96.6%
  • DeepSWE In use 113 original long-horizon engineering tasks across 91 repos and five languages, written from scratch. Hand-written verifiers accept alternative solutions. GPT-6 Astra 74.1%
  • SciCode In use Scientist-curated research coding problems split into subproblems across natural sciences. Code checked against test cases. Seed2.1 Pro 59.8%
  • Codeforces rating In use Model solves recent Codeforces contest problems. An Elo-style rating is computed from simulated contest performance. DeepSeek V4.1 Flash 3,471

Data audit

8 of 9 checks pass, 1 warning. Look first at the Task access check. We don’t know whether the tasks are public. Every benchmark

Catalog record 1 warning 1 pass
  • Warning Task access We don’t know whether the tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 3 pass 2 notes
  • Pass Independent runs Scores from the benchmark’s paper (6 entries).
  • Pass Models scored 6 models we track.
  • Pass Newest score The newest is from August 2026, for Qwen3.8 Flash.
  • Note Frontier models None of the top 10 has a score. The strongest tested is Claude Opus 5 (#11).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on github.com: 6 rows for models we track 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