Get started

Android native · Agents

Design Arena's Elo for android native · agents, from pairwise human preference votes, as OpenRouter carries it. 41 models ranked, as carried in OpenRouter's catalog.

retrieved sha256 · 0bb53a77713fwww.designarena.ai ↗

Most capable1389Claude Opus 5.5Anthropic
Best value2200MiniMax M31155 at $0.525 blended per M
Newest ranked1157Kimi K3.1 Prolisted 23 Oct 2026

5 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$10.0095611731389Blended list price per M tokens, log scaleEloClaude Opus 5.5: 1389 · $8.00 blended · Anthropic$8.00138905 Claude Opus 5.51389 · $8.00 blended · AnthropicMuse Spark 1.3: 1300 · $2.00 blended · Meta$2.00130004 Muse Spark 1.31300 · $2.00 blended · MetaGrok 4.6: 1279 · $3.00 blended · SpaceXAI$3.001279Grok 4.61279 · $3.00 blended · SpaceXAIClaude Fable 5: 1276 · $20.00 blended · Anthropic$20.001276Claude Fable 51276 · $20.00 blended · AnthropicGLM 5V Turbo: 1267 · $1.90 blended · Z.ai$1.90126703 GLM 5V Turbo1267 · $1.90 blended · Z.aiClaude Sonnet 5.5: 1260 · $4.00 blended · Anthropic$4.001260Claude Sonnet 5.51260 · $4.00 blended · AnthropicKimi K3: 1255 · $4.144 blended · MoonshotAI$4.1441255Kimi K31255 · $4.144 blended · MoonshotAIClaude Opus 5: 1245 · $10.00 blended · Anthropic$10.001245Claude Opus 51245 · $10.00 blended · AnthropicGemini 3.7 Flash: 1232 · $1.50 blended · Google$1.50123202 Gemini 3.7 Flash1232 · $1.50 blended · GoogleGrok 4.5: 1227 · $3.00 blended · SpaceXAI$3.001227Grok 4.51227 · $3.00 blended · SpaceXAIClaude Opus 4.8: 1222 · $10.00 blended · Anthropic$10.001222Claude Opus 4.81222 · $10.00 blended · AnthropicClaude Sonnet 5: 1210 · $4.00 blended · Anthropic$4.001210Claude Sonnet 51210 · $4.00 blended · AnthropicMuse Spark 1.2: 1209 · $2.00 blended · Meta$2.001209Muse Spark 1.21209 · $2.00 blended · MetaGPT-6.1 Sol: 1194 · $4.00 blended · OpenAI$4.001194GPT-6.1 Sol1194 · $4.00 blended · OpenAIGemini 3.6 Flash: 1193 · $1.50 blended · Google$1.501193Gemini 3.6 Flash1193 · $1.50 blended · GoogleClaude Sonnet 4.6: 1182 · $6.00 blended · Anthropic$6.001182Claude Sonnet 4.61182 · $6.00 blended · AnthropicClaude Opus 4.6: 1180 · $10.00 blended · Anthropic$10.001180Claude Opus 4.61180 · $10.00 blended · AnthropicGPT-5.2-Codex: 1176 · $4.813 blended · OpenAI$4.8131176GPT-5.2-Codex1176 · $4.813 blended · OpenAIGemini 3.5 Flash: 1174 · $3.375 blended · Google$3.3751174Gemini 3.5 Flash1174 · $3.375 blended · GoogleGLM 5.1: 1174 · $2.15 blended · Z.ai$2.151174GLM 5.11174 · $2.15 blended · Z.aiMuse Spark 1.1: 1172 · $2.00 blended · Meta$2.001172Muse Spark 1.11172 · $2.00 blended · MetaGLM 5.2: 1170 · $1.705 blended · Z.ai$1.7051170GLM 5.21170 · $1.705 blended · Z.aiClaude Opus 4.5: 1162 · $10.00 blended · Anthropic$10.001162Claude Opus 4.51162 · $10.00 blended · AnthropicKimi K3.1 Pro: 1157 · $2.281 blended · MoonshotAI$2.2811157Kimi K3.1 Pro1157 · $2.281 blended · MoonshotAIMiniMax M3: 1155 · $0.525 blended · MiniMax$0.525115501 MiniMax M31155 · $0.525 blended · MiniMaxGPT-5.5: 1153 · $11.25 blended · OpenAI$11.251153GPT-5.51153 · $11.25 blended · OpenAIKimi K2.6: 1151 · $1.34 blended · MoonshotAI$1.341151Kimi K2.61151 · $1.34 blended · MoonshotAIQwen3.7 Max: 1147 · $2.213 blended · Qwen$2.2131147Qwen3.7 Max1147 · $2.213 blended · QwenGLM 5: 1135 · $1.55 blended · Z.ai$1.551135GLM 51135 · $1.55 blended · Z.aiKimi K2.7 Code: 1129 · $1.463 blended · MoonshotAI$1.4631129Kimi K2.7 Code1129 · $1.463 blended · MoonshotAIKimi K2.5: 1105 · $1.00 blended · MoonshotAI$1.001105Kimi K2.51105 · $1.00 blended · MoonshotAIGLM 4.7: 1097 · $0.738 blended · Z.ai$0.7381097GLM 4.71097 · $0.738 blended · Z.aiGemini 3.8 Flash: 1087 · $1.65 blended · Google$1.651087Gemini 3.8 Flash1087 · $1.65 blended · GoogleGPT-5.3-Codex: 1084 · $4.813 blended · OpenAI$4.8131084GPT-5.3-Codex1084 · $4.813 blended · OpenAIGLM 4.6: 1072 · $0.875 blended · Z.ai$0.8751072GLM 4.61072 · $0.875 blended · Z.aiGPT-5.2: 1046 · $4.813 blended · OpenAI$4.8131046GPT-5.21046 · $4.813 blended · OpenAIGrok 4.20: 1036 · $1.563 blended · SpaceXAI$1.5631036Grok 4.201036 · $1.563 blended · SpaceXAIGemini 3.1 Pro Preview: 1035 · $4.50 blended · Google$4.501035Gemini 3.1 Pro Preview1035 · $4.50 blended · GoogleGemini 3 Flash Preview: 1034 · $1.125 blended · Google$1.1251034Gemini 3 Flash Preview1034 · $1.125 blended · GoogleGPT-5.4: 1021 · $5.625 blended · OpenAI$5.6251021GPT-5.41021 · $5.625 blended · OpenAIGrok 4.3: 956 · $1.563 blended · SpaceXAI$1.563956Grok 4.3956 · $1.563 blended · SpaceXAI
  1. 01MiniMax M31155 · $0.525 blended
  2. 02Gemini 3.7 Flash1232 · $1.50 blended · +77 at 2.9x
  3. 03GLM 5V Turbo1267 · $1.90 blended · +35 at 1.3x
  4. 04Muse Spark 1.31300 · $2.00 blended · +33 at 1.1x
  5. 05Claude Opus 5.51389 · $8.00 blended · +89 at 4.0x

41 of 41 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. 7 of 41 priced models sit in the attractive quadrant: at or above the median Elo of 1172 and at or below the median blended price of $2.281. The tinted corner is that quadrant. 41 of the 41 crosses match the selection below and are drawn white.

41 of 41.

Design Arena · Elo
PositionModelAuthorEloWin rateSource rankIn / out per MContextListed
01Claude Opus 5.5Anthropic138974.2%1$4.00 / $20.001M
02Muse Spark 1.3Meta130065.3%2$1.25 / $4.251.05M
03Grok 4.6SpaceXAI127961.0%3$2.00 / $6.00500k
04Claude Fable 5Anthropic127665.3%4$10.00 / $50.001M
05GLM 5V TurboZ.ai126754.8%5$1.20 / $4.00203k
06Claude Sonnet 5.5Anthropic126057.8%6$2.00 / $10.001M
07Kimi K3MoonshotAI125557.1%7$0.192 / $16.001.05M
08Claude Opus 5Anthropic124555.7%8$5.00 / $25.001M
09Gemini 3.7 FlashGoogle123253.2%9$0.75 / $3.751.05M
10Grok 4.5SpaceXAI122763.1%10$2.00 / $6.00500k
11Claude Opus 4.8Anthropic122255.9%11$5.00 / $25.001M
12Claude Sonnet 5Anthropic121053.0%12$2.00 / $10.001M
13Muse Spark 1.2Meta120952.3%13$1.25 / $4.251.05M
14GPT-6.1 SolOpenAI119449.0%15$2.00 / $10.001.05M
15Gemini 3.6 FlashGoogle119353.4%16$0.75 / $3.751.05M
16Claude Sonnet 4.6Anthropic118261.8%17$3.00 / $15.001M
17Claude Opus 4.6Anthropic118061.0%18$5.00 / $25.001M
18GPT-5.2-CodexOpenAI117647.5%19$1.75 / $14.00400k
19Gemini 3.5 FlashGoogle117457.2%20$1.50 / $9.001.05M
20GLM 5.1Z.ai117451.9%21$1.40 / $4.40205k
21Muse Spark 1.1Meta117248.0%22$1.25 / $4.251.05M
22GLM 5.2Z.ai117052.8%23$0.007 / $6.801.05M
23Claude Opus 4.5Anthropic116265.5%24$5.00 / $25.00200k
24Kimi K3.1 ProMoonshotAI115743.8%25$0.24 / $8.401.05M
25MiniMax M3MiniMax115544.9%26$0.30 / $1.201.05M
26GPT-5.5OpenAI115350.8%27$5.00 / $30.001.05M
27Kimi K2.6MoonshotAI115149.2%28$0.65 / $3.41262k
28Qwen3.7 MaxQwen114747.5%29$1.48 / $4.421M
29GLM 5Z.ai113555.0%30$1.00 / $3.20205k
30Kimi K2.7 CodeMoonshotAI112946.0%31$0.70 / $3.75262k
31Kimi K2.5MoonshotAI110557.8%33$0.50 / $2.50262k
32GLM 4.7Z.ai109755.8%34$0.40 / $1.75205k
33Gemini 3.8 FlashGoogle108734.3%35$0.825 / $4.131.05M
34GPT-5.3-CodexOpenAI108435.2%36$1.75 / $14.00400k
35GLM 4.6Z.ai107252.8%37$0.50 / $2.00205k
36GPT-5.2OpenAI104649.1%38$1.75 / $14.00400k
37Grok 4.20SpaceXAI103635.0%39$1.25 / $2.502M
38Gemini 3.1 Pro PreviewGoogle103541.1%40$2.00 / $12.001.05M
39Gemini 3 Flash PreviewGoogle103447.7%41$0.50 / $3.001.05M
40GPT-5.4OpenAI102147.7%42$2.50 / $15.001.05M
41Grok 4.3SpaceXAI95623.0%43$1.25 / $2.501M

The field and the frontier above are the whole board and do not narrow with this table: a frontier drawn through 41 of 41 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.