Get started

Agentic game development · Agents

Design Arena's Elo for agentic game development · agents, from pairwise human preference votes, as OpenRouter carries it. 25 models ranked, as carried in OpenRouter's catalog.

retrieved sha256 · 0bb53a77713fwww.designarena.ai ↗

Most capable1266Gemini 4 ArgonGoogle
Best value2118MiniMax M31112 at $0.525 blended per M
Newest ranked1266Gemini 4 Argonlisted 13 Oct 2026

4 models sit on the frontier: nothing scores higher for less.

  • the attractive quadrant
  • the frontier
  • a priced model, in its maker's tone
  • on the frontier, named
The attractive quadrant$1.00$10.0095411101266Blended list price per M tokens, log scaleEloGemini 4 Argon: 1266 · $4.00 blended · Google$4.00126604 Gemini 4 Argon1266 · $4.00 blended · GoogleClaude Fable 5: 1231 · $20.00 blended · Anthropic$20.001231Claude Fable 51231 · $20.00 blended · AnthropicClaude Opus 5: 1202 · $10.00 blended · Anthropic$10.001202Claude Opus 51202 · $10.00 blended · AnthropicKimi K3: 1201 · $4.144 blended · MoonshotAI$4.1441201Kimi K31201 · $4.144 blended · MoonshotAIClaude Opus 4.8: 1182 · $10.00 blended · Anthropic$10.001182Claude Opus 4.81182 · $10.00 blended · AnthropicGrok 4.5: 1181 · $3.00 blended · SpaceXAI$3.00118103 Grok 4.51181 · $3.00 blended · SpaceXAIGemini 3.7 Flash: 1180 · $1.50 blended · Google$1.50118002 Gemini 3.7 Flash1180 · $1.50 blended · GoogleClaude Sonnet 5: 1169 · $4.00 blended · Anthropic$4.001169Claude Sonnet 51169 · $4.00 blended · AnthropicGLM 5.2: 1156 · $1.705 blended · Z.ai$1.7051156GLM 5.21156 · $1.705 blended · Z.aiGrok 4.6: 1155 · $3.00 blended · SpaceXAI$3.001155Grok 4.61155 · $3.00 blended · SpaceXAIClaude Opus 4.6: 1143 · $10.00 blended · Anthropic$10.001143Claude Opus 4.61143 · $10.00 blended · AnthropicMuse Spark 1.1: 1130 · $2.00 blended · Meta$2.001130Muse Spark 1.11130 · $2.00 blended · MetaGemini 3.5 Flash: 1126 · $3.375 blended · Google$3.3751126Gemini 3.5 Flash1126 · $3.375 blended · GoogleMuse Spark 1.2: 1125 · $2.00 blended · Meta$2.001125Muse Spark 1.21125 · $2.00 blended · MetaGPT-5.5: 1124 · $11.25 blended · OpenAI$11.251124GPT-5.51124 · $11.25 blended · OpenAIGemini 3.6 Flash: 1123 · $1.50 blended · Google$1.501123Gemini 3.6 Flash1123 · $1.50 blended · GoogleClaude Sonnet 4.6: 1117 · $6.00 blended · Anthropic$6.001117Claude Sonnet 4.61117 · $6.00 blended · AnthropicMiniMax M3: 1112 · $0.525 blended · MiniMax$0.525111201 MiniMax M31112 · $0.525 blended · MiniMaxGLM 5.1: 1108 · $2.15 blended · Z.ai$2.151108GLM 5.11108 · $2.15 blended · Z.aiQwen3.7 Max: 1104 · $2.213 blended · Qwen$2.2131104Qwen3.7 Max1104 · $2.213 blended · QwenKimi K2.6: 1087 · $1.34 blended · MoonshotAI$1.341087Kimi K2.61087 · $1.34 blended · MoonshotAIKimi K2.7 Code: 1076 · $1.463 blended · MoonshotAI$1.4631076Kimi K2.7 Code1076 · $1.463 blended · MoonshotAIGemini 3.1 Pro Preview: 1060 · $4.50 blended · Google$4.501060Gemini 3.1 Pro Preview1060 · $4.50 blended · GoogleGLM 5V Turbo: 1048 · $1.90 blended · Z.ai$1.901048GLM 5V Turbo1048 · $1.90 blended · Z.aiGrok 4.3: 954 · $1.563 blended · SpaceXAI$1.563954Grok 4.3954 · $1.563 blended · SpaceXAI
  1. 01MiniMax M31112 · $0.525 blended
  2. 02Gemini 3.7 Flash1180 · $1.50 blended · +68 at 2.9x
  3. 03Grok 4.51181 · $3.00 blended · +1 at 2.0x
  4. 04Gemini 4 Argon1266 · $4.00 blended · +85 at 1.3x

25 of 25 ranked models carry a list price; 0 are free or unlisted and not in the field. Price is the blended OpenRouter rate, three input tokens to one output, on a logarithmic axis. Each dot is filled in its maker's tone, as the columns above are. The numbered and named dots are the frontier, cheapest first; each step after the first names what it adds over the step before, and at what multiple of the blended price. A hover over any dot names it and its two figures, a click opens the model, and a model named in the picker above keeps its label on. 5 of 25 priced models sit in the attractive quadrant: at or above the median Elo of 1126 and at or below the median blended price of $3.00. The tinted corner is that quadrant. 5 of the 25 crosses match the selection below and are drawn white.

5 of 25.

Design Arena · Elo
PositionModelAuthorEloWin rateSource rankIn / out per MContextListed
01Gemini 4 ArgonGoogle126651.1%1$2.00 / $10.001.05M
07Gemini 3.7 FlashGoogle118052.0%7$0.75 / $3.751.05M
13Gemini 3.5 FlashGoogle112655.3%13$1.50 / $9.001.05M
16Gemini 3.6 FlashGoogle112347.3%16$0.75 / $3.751.05M
23Gemini 3.1 Pro PreviewGoogle106043.9%23$2.00 / $12.001.05M

The field and the frontier above are the whole board and do not narrow with this table: a frontier drawn through 5 of 25 models would say nothing beats them on both while something in the same board does. The models this table holds are drawn white in the figures instead.

What the score tells you

An Elo from pairwise human preference votes: a higher figure means the model's output was preferred more often against the field it was matched with. Win rate is the share of matchups won. Both shift as new models enter the arena, neither is a measure of correctness, and the source's own rank counts models this catalog does not carry.

Sources

Source: OpenRouter (openrouter.ai/models), as of 17 Nov 16:29 UTC. sha256 0bb53a77713f over 853,270 bytes.
Catalog: OpenRouter's public model list.
Intelligence, coding and agentic indexes: Artificial Analysis, as published by OpenRouter.
Elo, win rate and rank: Design Arena, as published by OpenRouter.
Vitna ran none of these evaluations and adjusted none of these figures.