Undominated.ai

Alternatives to GPT-5 Mini

OpenAI LMArena 1373.2 $0.688/M on a balanced workload prices as of 2026-10-06

2 models are higher-scoring or cheaper, with neither dimension worse.

Of the 144 models carrying both an independent score and a published price at the same delivery mode, 36 score at least as high as GPT-5 Mini and cost no more under at least one listed workload, with at least one of those dimensions improved. 2 match or beat its recorded context, maximum output, input modes, tool use and extended reasoning. 34 have a recorded capability loss, named on every row.

Dominance is not a property of a model. It is a property of a model, a capability lens and a workload, and all three are named on every row. Scores are LMArena Elo, used under CC BY 4.0 from the official dataset; prices are blended per million tokens. A higher score is not a drop-in replacement.

This page is built only while a candidate improves score or price without worsening the other under a listed workload. When that comparison no longer holds, it is not generated.

Where it holds

WorkloadGPT-5 Mini $/MRecorded capabilities preservedRecorded capability losses
Balanced$0.688232
Summarise$0.273222
Chat$0.883234
Code gen$1.28234
Agentic$0.379229

Workload mixes are defined on the methodology page. Cached input is priced at the cached rate, and reasoning tokens at the worse of the reasoning and output rates.

What you would be replacing

Intelligence1373.2
Context window400K
Max output128K
Input modestext, image, file
Tool useyes
Extended reasoningyes

Preserving recorded capabilities requires matching or beating these fields. Full record for GPT-5 Mini, or check GPT-5 Mini against the whole catalogue.

Recorded capabilities preserved 2

Each option has a higher score or a lower price, with neither dimension worse under the listed workloads. It matches or beats GPT-5 Mini on recorded context, maximum output, input modes, tool use and extended reasoning.

#ModelIntelligence$/MHolds under
1GPT-5.6 Luna OpenAI1431 +57.8$0.450
  • Balanced −35%
  • Summarise −27%
  • Chat −38%
  • Code gen −39%
  • Agentic −36%
2GPT-6 Luna OpenAI1391.5 +18.3$0.200
  • Balanced −71%
  • Summarise −65%
  • Chat −74%
  • Code gen −74%
  • Agentic −72%

Alternatives with recorded capability losses 34

These improve score or price without worsening the other under the listed workloads, but reduce at least one recorded capability. Read the last column before switching.

ModelIntelligence$/MHolds underRecorded capability loss
MiMo-V2.6-Pro Xiaomi1491 +117.8$0.540
  • Balanced −21%
  • Chat −46%
  • Code gen −49%
  • Agentic −36%
  • no file input
GLM 5.3 Flash Z.ai1469.6 +96.4$0.238
  • Balanced −65%
  • Summarise −51%
  • Chat −71%
  • Code gen −73%
  • Agentic −65%
  • no file input
DeepSeek V4.1 Flash DeepSeek1462.5 +89.3$0.182
  • Balanced −73%
  • Summarise −76%
  • Chat −72%
  • Code gen −72%
  • Agentic −72%
  • no file input
MiMo-V2.5-Pro Xiaomi1465 +91.8$0.544
  • Balanced −21%
  • Chat −46%
  • Code gen −48%
  • Agentic −36%
  • no image, file input
MiMo-V2.6-Flash Xiaomi1456.4 +83.2$0.175
  • Balanced −75%
  • Summarise −61%
  • Chat −82%
  • Code gen −83%
  • Agentic −79%
  • no file input
Qwen3.7 Plus Qwen1454.6 +81.4$0.560
  • Balanced −19%
  • Chat −29%
  • Code gen −32%
  • Agentic −18%
  • no file input
Gemma 4 31B Google1443.4 +70.2$0.205
  • Balanced −70%
  • Summarise −44%
  • Chat −72%
  • Code gen −77%
  • Agentic −53%
  • 400K → 262K context
  • 128K → 16K max output
  • no file input
Hy3 Tencent1440.7 +67.5$0.231
  • Balanced −66%
  • Summarise −55%
  • Chat −70%
  • Code gen −72%
  • Agentic −65%
  • 400K → 262K context
  • no image, file input
Qwen3.8 27B Qwen1440.7 +67.5$0.562
  • Balanced −18%
  • Summarise −33%
  • Chat −4%
  • Code gen −5%
  • Agentic −2%
  • no file input
Gemma 4 26B A4B Google1433.9 +60.7$0.127
  • Balanced −82%
  • Summarise −75%
  • Chat −83%
  • Code gen −84%
  • Agentic −79%
  • 400K → 262K context
  • no file input
DeepSeek V4 Flash 0423 DeepSeek1432.1 +58.9$0.113
  • Balanced −84%
  • Summarise −73%
  • Chat −88%
  • Code gen −89%
  • Agentic −84%
  • no image, file input
MiniMax M3 MiniMax1432.1 +58.9$0.485
  • Balanced −29%
  • Summarise −6%
  • Chat −39%
  • Code gen −41%
  • Agentic −29%
  • no file input
MiMo-V2.5 Xiaomi1427.5 +54.3$0.175
  • Balanced −75%
  • Summarise −61%
  • Chat −82%
  • Code gen −83%
  • Agentic −79%
  • no file input
DeepSeek V3.2 DeepSeek1424.5 +51.3$0.290
  • Balanced −58%
  • Summarise −16%
  • Chat −70%
  • Code gen −75%
  • Agentic −47%
  • 400K → 164K context
  • 128K → 66K max output
  • no image, file input
DeepSeek V3.2 Exp DeepSeek1422.4 +49.2$0.305
  • Balanced −56%
  • Chat −63%
  • Code gen −72%
  • Agentic −23%
  • 400K → 164K context
  • no image, file input
Qwen3 235B A22B Instruct 2507 Qwen1419.3 +46.1$0.205
  • Balanced −70%
  • Summarise −59%
  • Chat −69%
  • Code gen −71%
  • Agentic −58%
  • 400K → 262K context
  • 128K → 16K max output
  • no image, file input
  • no extended reasoning
Qwen3 VL 235B A22B Instruct Qwen1420.1 +46.9$0.370
  • Balanced −46%
  • Summarise −24%
  • Chat −50%
  • Code gen −53%
  • Agentic −34%
  • 400K → 262K context
  • 128K → 33K max output
  • no file input
  • no extended reasoning
Qwen3 Next 80B A3B Instruct Qwen1417.3 +44.1$0.268
  • Balanced −61%
  • Summarise −52%
  • Chat −58%
  • Code gen −60%
  • Agentic −47%
  • 400K → 262K context
  • no image, file input
  • no extended reasoning
DeepSeek V3.1 DeepSeek1419.7 +46.5$0.453
  • Balanced −34%
  • Chat −36%
  • Code gen −45%
  • 400K → 164K context
  • 128K → 33K max output
  • no image, file input
DeepSeek V3.1 Terminus DeepSeek1417.1 +43.9$0.453
  • Balanced −34%
  • Chat −36%
  • Code gen −45%
  • 400K → 164K context
  • no image, file input
Gemini 3.1 Flash Lite Preview Google1415.7 +42.5$0.563
  • Balanced −18%
  • Summarise −9%
  • Chat −23%
  • Code gen −23%
  • Agentic −20%
  • 128K → 66K max output
Inkling Small Thinking Machines1413.7 +40.5$0.638
  • Balanced −7%
  • Chat −27%
  • Code gen −32%
  • Agentic −6%
  • no file input
Qwen3.5-27B Qwen1408.9 +35.7$0.536
  • Balanced −22%
  • Summarise −4%
  • Chat −16%
  • Code gen −21%
  • 400K → 262K context
  • 128K → 66K max output
  • no file input
Step 3.5 Flash StepFun1403 +29.8$0.150
  • Balanced −78%
  • Summarise −60%
  • Chat −80%
  • Code gen −83%
  • Agentic −66%
  • 400K → 262K context
  • 128K → 66K max output
  • no image, file input
MiniMax M2.7 MiniMax1404.6 +31.4$0.525
  • Balanced −24%
  • Chat −33%
  • Code gen −36%
  • Agentic −23%
  • 400K → 205K context
  • no image, file input
Qwen3.5-Flash Qwen1397 +23.8$0.114
  • Balanced −83%
  • Summarise −73%
  • Chat −84%
  • Code gen −86%
  • Agentic −75%
  • 128K → 66K max output
  • no file input
Qwen3.5-35B-A3B Qwen1394.4 +21.2$0.355
  • Balanced −48%
  • Summarise −42%
  • Chat −48%
  • Code gen −49%
  • Agentic −43%
  • 400K → 262K context
  • 128K → 33K max output
  • no file input
Qwen3 30B A3B Instruct 2507 Qwen1384.1 +10.9$0.143
  • Balanced −79%
  • Summarise −63%
  • Chat −80%
  • Code gen −83%
  • Agentic −68%
  • 400K → 262K context
  • no image, file input
  • no extended reasoning
GLM 4.5 Air Z.ai1383.9 +10.7$0.310
  • Balanced −55%
  • Summarise −50%
  • Chat −56%
  • Code gen −57%
  • Agentic −54%
  • 400K → 131K context
  • 128K → 98K max output
  • no image, file input
Solar Pro 4 Upstage1386.2 +13$0.525
  • Balanced −24%
  • Chat −33%
  • Code gen −36%
  • Agentic −23%
  • no image, file input
GLM 4.6V Z.ai1376.8 tie$0.450
  • Balanced −35%
  • Summarise −5%
  • Chat −47%
  • Code gen −50%
  • Agentic −36%
  • 400K → 131K context
  • 128K → 33K max output
  • no file input
DeepSeek V3 0324 DeepSeek1375 tie$0.438
  • Balanced −36%
  • Chat −38%
  • Code gen −45%
  • Agentic −4%
  • 400K → 164K context
  • 128K → 115K max output
  • no image, file input
  • no extended reasoning
GLM 4.6 Z.ai1439.9 +66.7$0.853 chat
  • Chat −3%
  • Code gen −7%
  • 400K → 205K context
  • 128K → 16K max output
  • no image, file input
GLM 4.7 Z.ai1435.5 +62.3$0.844 chat
  • Chat −4%
  • Code gen −8%
  • 400K → 205K context
  • no image, file input

What this compares, and what it leaves out

  • Quality is LMArena Elo, used under CC BY 4.0 from the official dataset. The 95% confidence interval on a difference between two scores is about ±10.47 points, so a gap smaller than that is marked tie rather than an improvement — see significance bands.
  • 204 further models at this delivery mode carry a price but no independent score. They are absent from the comparison in both directions — unrated is not a zero, and an unmeasured model is neither an alternative nor a worse buy.
  • Retired models are never offered as an alternative, and a model only competes against its own delivery mode: batch trades latency for price, so it is not a like-for-like swap.
  • Nothing here measures latency, throughput, rate limits or how a model behaves on your prompts. Two models with the same index score are not interchangeable.
  • Models with no higher-scoring or cheaper alternative that is no worse on the other axis are on the value frontier. Models with such an alternative are listed here.
Evidence & Ask
Markdown for LLMs