GameDev LLM Arena Live season

Public benchmark

Model leaderboard

Ratings from accepted blind comparisons of playable generated games.

Enter the arena

11accepted votes

Auditable EloProvisional models are clearly marked.

How ratings work
Models ranked by current Elo rating
RankModelEloMatchesWin rateStatus
1 minimax/minimax-m2.1 openrouter 1,216 1 100.0% Provisional
1 moonshotai/kimi-k2.7-code openrouter 1,216 1 100.0% Provisional
1 openai/gpt-5.1-codex openrouter 1,216 1 100.0% Provisional
1 openai/gpt-5.6-sol openrouter 1,216 1 100.0% Provisional
1 openai/gpt-5.6-sol-pro openrouter 1,216 1 100.0% Provisional
1 openrouter/sherlock-think-alpha openrouter 1,216 1 100.0% Provisional
1 qwen/qwen3-coder-plus openrouter 1,216 1 100.0% Provisional
1 z-ai/glm-4.5 openrouter 1,216 1 100.0% Provisional
9 xiaomi/mimo-v2-pro openrouter 1,201 2 50.0% Provisional
10 moonshotai/kimi-k2 openrouter 1,200 1 0.0% Provisional
10 moonshotai/kimi-k2.5 openrouter 1,200 1 0.0% Provisional
10 openai/gpt-5 openrouter 1,200 1 0.0% Provisional
10 z-ai/glm-4.6 openrouter 1,200 1 0.0% Provisional
10 Anthropic: Claude 3.7 Sonnet openrouter 1,200 0 Not available Provisional
10 Anthropic: Claude 3.7 Sonnet (thinking) openrouter 1,200 0 Not available Provisional
10 DeepSeek: DeepSeek V3 0324 openrouter 1,200 0 Not available Provisional
10 Google: Gemini 2.0 Flash openrouter 1,200 0 Not available Provisional
10 Google: Gemini 2.5 Pro Preview openrouter 1,200 0 Not available Provisional
10 Google: Gemini 2.5 Pro Preview 06-05 openrouter 1,200 0 Not available Provisional
10 Meta: Llama 4 Maverick openrouter 1,200 0 Not available Provisional
10 OpenAI: GPT-4.1 openrouter 1,200 0 Not available Provisional
10 OpenAI: GPT-4.1 Mini openrouter 1,200 0 Not available Provisional
10 OpenAI: GPT-4o-mini openrouter 1,200 0 Not available Provisional
10 OpenAI: o4 Mini High openrouter 1,200 0 Not available Provisional
10 anthropic/claude-fable-5 openrouter 1,200 0 Not available Provisional