Undominated.ai

Alternatives to Step 5 Preview

StepFun LMArena 1440 $1.43/M on a balanced workload prices as of 2026-10-21, except 9 under confirmation

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

Of the 142 models carrying both an independent score and a published price at the same delivery mode, 15 score at least as high as Step 5 Preview and cost no more under at least one listed workload, with at least one of those dimensions improved. 4 match or beat its recorded context, maximum output, input modes, tool use and extended reasoning. 11 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 lists models only while a candidate improves score or price without worsening the other under a listed workload. When that comparison no longer holds, this address stays and says so.

Where it holds

WorkloadStep 5 Preview $/MRecorded capabilities preservedRecorded capability losses
Balanced$1.43311
Summarise$0.829411
Chat$1.4138
Code gen$1.9537
Agentic$0.720311

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

Intelligence1440
Context window1M
Max output64K
Input modestext, image, video
Tool useyes
Extended reasoningyes

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

Recorded capabilities preserved 4

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

#ModelIntelligence$/MHolds under
1MiMo-V2.6-Pro Xiaomi1490.3 +50.3$0.540 Price under confirmation since Oct 21, 2026
  • Balanced −62%
  • Summarise −60%
  • Chat −66%
  • Code gen −66%
  • Agentic −66%
2GLM 5.3 Flash Z.ai1471 +31$0.238 Price under confirmation since Oct 12, 2026
  • Balanced −83%
  • Summarise −84%
  • Chat −82%
  • Code gen −82%
  • Agentic −82%
3MiMo-V2.6-Flash Xiaomi1455.2 +15.2$0.168 Price under confirmation since Oct 21, 2026
  • Balanced −88%
  • Summarise −88%
  • Chat −89%
  • Code gen −89%
  • Agentic −89%
4Gemini 3.6 Flash Google1479.5 +39.5$0.708 summarise
  • Summarise −15%

Alternatives with recorded capability losses 11

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
GLM 5.3 Z.ai1471.4 +31.4$1.00 Price under confirmation since Oct 12, 2026
  • Balanced −30%
  • Summarise −58%
  • Agentic −10%
  • no image, video input
GLM 5.2 Z.ai1470 +30$1.12 Price under confirmation since Oct 12, 2026
  • Balanced −21%
  • Summarise −49%
  • Agentic −4%
  • no image, video input
MiMo-V2.5-Pro Xiaomi1465.1 +25.1$0.544 Price under confirmation since Oct 21, 2026
  • Balanced −62%
  • Summarise −60%
  • Chat −66%
  • Code gen −66%
  • Agentic −66%
  • no image, video input
DeepSeek V4.1 Flash DeepSeek1462.5 +22.5$0.182 Price under confirmation since Oct 12, 2026
  • Balanced −87%
  • Summarise −92%
  • Chat −82%
  • Code gen −82%
  • Agentic −85%
  • no video input
Qwen3.7 Plus Qwen1454.6 +14.6$0.560
  • Balanced −61%
  • Summarise −64%
  • Chat −56%
  • Code gen −55%
  • Agentic −57%
  • 1M → 1M context
  • no video input
Kimi K2.6 Moonshot AI1455.4 +15.4$0.825
  • Balanced −42%
  • Summarise −51%
  • Chat −34%
  • Code gen −32%
  • Agentic −40%
  • 1M → 262K context
  • no video input
Gemma 4 31B Google1442.9 tie$0.205
  • Balanced −86%
  • Summarise −82%
  • Chat −83%
  • Code gen −85%
  • Agentic −75%
  • 1M → 262K context
  • 64K → 16K max output
GLM 5 Z.ai1446.1 tie$1.09
  • Balanced −24%
  • Summarise −25%
  • Chat −19%
  • Code gen −19%
  • Agentic −17%
  • 1M → 205K context
  • no image, video input
DeepSeek V4 Pro 0423 DeepSeek1445.4 tie$1.21 Price under confirmation since Oct 21, 2026
  • Balanced −15%
  • Summarise −39%
  • Agentic −3%
  • no image, video input
Kimi K2.5 Moonshot AI1445 tie$1.20
  • Balanced −16%
  • Summarise −30%
  • Chat same price
  • Agentic −8%
  • 1M → 262K context
  • no video input
Qwen3.8 27B Qwen1440.6 tie$0.956 Price under confirmation since Oct 21, 2026
  • Balanced −33%
  • Summarise −48%
  • Chat −17%
  • Code gen −14%
  • Agentic −25%
  • 1M → 1M context

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 ±10.36 points, so a gap smaller than that is marked tie rather than an improvement — see significance bands.
  • 203 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