Undominated.ai

Alternatives to Mistral Medium 3.5

Mistral LMArena 1421.2 $3.00/M on a balanced workload prices as of 2026-10-06

5 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, 48 score at least as high as Mistral Medium 3.5 and cost no more under at least one listed workload, with at least one of those dimensions improved. 5 match or beat its recorded context, maximum output, input modes, tool use and extended reasoning. 43 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

WorkloadMistral Medium 3.5 $/MRecorded capabilities preservedRecorded capability losses
Balanced$3.00536
Summarise$1.80540
Chat$3.90536
Code gen$5.10536
Agentic$2.40543

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

Intelligence1421.2
Context window262K
Max output209K
Input modestext, image, file
Tool useyes
Extended reasoningyes

Preserving recorded capabilities requires matching or beating these fields. Full record for Mistral Medium 3.5, or check Mistral Medium 3.5 against the whole catalogue.

Recorded capabilities preserved 5

Each option has a higher score or a lower price, with neither dimension worse under the listed workloads. It matches or beats Mistral Medium 3.5 on recorded context, maximum output, input modes, tool use and extended reasoning.

#ModelIntelligence$/MHolds under
1Muse Spark 1.3 Meta1489.9 +68.7$2.00
  • Balanced −33%
  • Summarise −40%
  • Chat −46%
  • Code gen −42%
  • Agentic −56%
2Muse Spark 1.2 Meta1483.2 +62$2.00
  • Balanced −33%
  • Summarise −40%
  • Chat −46%
  • Code gen −42%
  • Agentic −56%
3Muse Spark 1.1 Meta1479.2 +58$2.00
  • Balanced −33%
  • Summarise −40%
  • Chat −46%
  • Code gen −42%
  • Agentic −56%
4Grok 4.5 xAI1448.1 +26.9$3.00
  • Balanced same price
  • Summarise −5%
  • Chat −21%
  • Code gen −16%
  • Agentic −34%
5Grok 4.6 xAI1427.4 tie$3.00
  • Balanced same price
  • Summarise −2%
  • Chat −19%
  • Code gen −16%
  • Agentic −29%

Alternatives with recorded capability losses 43

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 +69.8$0.540
  • Balanced −82%
  • Summarise −82%
  • Chat −88%
  • Code gen −87%
  • Agentic −90%
  • 210K → 131K max output
  • no file input
Gemini 3.8 Flash Google1497 +75.8$3.00
  • Balanced same price
  • Summarise −21%
  • Chat −10%
  • Code gen −2%
  • Agentic −33%
  • 210K → 66K max output
Gemini 3.7 Flash Google1487 +65.8$3.00
  • Balanced same price
  • Summarise −21%
  • Chat −10%
  • Code gen −2%
  • Agentic −33%
  • 210K → 66K max output
Gemini 3.6 Flash Google1479.5 +58.3$1.50
  • Balanced −50%
  • Summarise −61%
  • Chat −55%
  • Code gen −51%
  • Agentic −67%
  • 210K → 66K max output
GLM 5.3 Flash Z.ai1469.6 +48.4$0.238
  • Balanced −92%
  • Summarise −93%
  • Chat −93%
  • Code gen −93%
  • Agentic −95%
  • no file input
GLM 5.3 Z.ai1471.4 +50.2$1.00
  • Balanced −67%
  • Summarise −81%
  • Chat −62%
  • Code gen −59%
  • Agentic −73%
  • no image, file input
GLM 5.2 Z.ai1470.4 +49.2$1.12
  • Balanced −63%
  • Summarise −77%
  • Chat −60%
  • Code gen −56%
  • Agentic −71%
  • 210K → 131K max output
  • no image, file input
MiMo-V2.5-Pro Xiaomi1465 +43.8$0.544
  • Balanced −82%
  • Summarise −81%
  • Chat −88%
  • Code gen −87%
  • Agentic −90%
  • 210K → 131K max output
  • no image, file input
DeepSeek V4.1 Flash DeepSeek1462.5 +41.3$0.182
  • Balanced −94%
  • Summarise −96%
  • Chat −94%
  • Code gen −93%
  • Agentic −96%
  • no file input
MiMo-V2.6-Flash Xiaomi1456.4 +35.2$0.175
  • Balanced −94%
  • Summarise −94%
  • Chat −96%
  • Code gen −96%
  • Agentic −97%
  • 210K → 131K max output
  • no file input
GLM 5.1 Z.ai1461.5 +40.3$1.83
  • Balanced −39%
  • Summarise −36%
  • Chat −48%
  • Code gen −48%
  • Agentic −49%
  • 262K → 205K context
  • 210K → 131K max output
  • no image, file input
Qwen3.7 Plus Qwen1454.6 +33.4$0.560
  • Balanced −81%
  • Summarise −84%
  • Chat −84%
  • Code gen −83%
  • Agentic −87%
  • 210K → 131K max output
  • no file input
Kimi K2.6 Moonshot AI1455.5 +34.3$0.961
  • Balanced −68%
  • Summarise −74%
  • Chat −71%
  • Code gen −68%
  • Agentic −77%
  • no file input
Gemma 4 31B Google1443.4 +22.2$0.205
  • Balanced −93%
  • Summarise −92%
  • Chat −94%
  • Code gen −94%
  • Agentic −93%
  • 210K → 16K max output
  • no file input
GLM 5 Z.ai1446.3 +25.1$1.09
  • Balanced −64%
  • Summarise −66%
  • Chat −71%
  • Code gen −69%
  • Agentic −75%
  • 262K → 205K context
  • 210K → 128K max output
  • no image, file input
Kimi K2.5 Moonshot AI1445.2 +24$1.20
  • Balanced −60%
  • Summarise −68%
  • Chat −64%
  • Code gen −61%
  • Agentic −72%
  • no file input
DeepSeek V4 Pro 0423 DeepSeek1445.2 +24$1.21
  • Balanced −60%
  • Summarise −72%
  • Chat −60%
  • Code gen −56%
  • Agentic −71%
  • no image, file input
Hy3 Tencent1440.7 +19.5$0.231
  • Balanced −92%
  • Summarise −93%
  • Chat −93%
  • Code gen −93%
  • Agentic −94%
  • 210K → 128K max output
  • no image, file input
Qwen3.6 Max Preview Qwen1446.8 +25.6$2.31
  • Balanced −23%
  • Summarise −29%
  • Chat −21%
  • Code gen −19%
  • Agentic −25%
  • 210K → 66K max output
  • no image, file input
Qwen3.8 27B Qwen1440.7 +19.5$0.562
  • Balanced −81%
  • Summarise −90%
  • Chat −78%
  • Code gen −76%
  • Agentic −85%
  • 210K → 131K max output
  • no file input
GLM 4.6 Z.ai1439.9 +18.7$0.760
  • Balanced −75%
  • Summarise −78%
  • Chat −78%
  • Code gen −77%
  • Agentic −83%
  • 262K → 205K context
  • 210K → 16K max output
  • no image, file input
Inkling Thinking Machines1441.8 +20.6$1.73
  • Balanced −43%
  • Summarise −51%
  • Chat −50%
  • Code gen −46%
  • Agentic −61%
  • no file input
Qwen3.5 397B A17B Qwen1438 +16.8$0.878
  • Balanced −71%
  • Summarise −73%
  • Chat −70%
  • Code gen −69%
  • Agentic −72%
  • 210K → 82K max output
  • no file input
Qwen3.6 Plus Qwen1436.7 +15.5$0.731
  • Balanced −76%
  • Summarise −77%
  • Chat −75%
  • Code gen −75%
  • Agentic −76%
  • 210K → 66K max output
  • no file input
Gemma 4 26B A4B Google1433.9 +12.7$0.127
  • Balanced −96%
  • Summarise −96%
  • Chat −96%
  • Code gen −96%
  • Agentic −97%
  • no file input
GLM 4.7 Z.ai1435.5 +14.3$0.738
  • Balanced −75%
  • Summarise −79%
  • Chat −78%
  • Code gen −77%
  • Agentic −83%
  • 262K → 205K context
  • 210K → 131K max output
  • no image, file input
Gemini 3.5 Flash Lite Google1434.7 +13.5$0.850
  • Balanced −72%
  • Summarise −81%
  • Chat −72%
  • Code gen −69%
  • Agentic −80%
  • 210K → 66K max output
DeepSeek V4 Flash 0423 DeepSeek1432.1 +10.9$0.113
  • Balanced −96%
  • Summarise −96%
  • Chat −97%
  • Code gen −97%
  • Agentic −97%
  • no image, file input
GLM 5V Turbo Z.ai1437.2 +16$1.90
  • Balanced −37%
  • Summarise −41%
  • Chat −48%
  • Code gen −45%
  • Agentic −56%
  • 262K → 203K context
  • 210K → 131K max output
  • no file input
MiniMax M3 MiniMax1432.1 +10.9$0.485
  • Balanced −84%
  • Summarise −86%
  • Chat −86%
  • Code gen −85%
  • Agentic −89%
  • no file input
GPT-5.6 Luna OpenAI1431 tie$0.450
  • Balanced −85%
  • Summarise −89%
  • Chat −86%
  • Code gen −85%
  • Agentic −90%
  • 210K → 128K max output
GLM 4.5 Z.ai1430.3 tie$1.00
  • Balanced −67%
  • Summarise −70%
  • Chat −72%
  • Code gen −70%
  • Agentic −77%
  • 262K → 131K context
  • 210K → 98K max output
  • no image, file input
MiMo-V2.5 Xiaomi1427.5 tie$0.175
  • Balanced −94%
  • Summarise −94%
  • Chat −96%
  • Code gen −96%
  • Agentic −97%
  • 210K → 131K max output
  • no file input
R1 0528 DeepSeek1427.9 tie$0.920
  • Balanced −69%
  • Summarise −68%
  • Chat −70%
  • Code gen −70%
  • Agentic −69%
  • 262K → 164K context
  • 210K → 33K max output
  • no image, file input
DeepSeek V3.2 DeepSeek1424.5 tie$0.290
  • Balanced −90%
  • Summarise −87%
  • Chat −93%
  • Code gen −94%
  • Agentic −92%
  • 262K → 164K context
  • 210K → 66K max output
  • no image, file input
DeepSeek V3.2 Exp DeepSeek1422.4 tie$0.305
  • Balanced −90%
  • Summarise −85%
  • Chat −92%
  • Code gen −93%
  • Agentic −88%
  • 262K → 164K context
  • 210K → 147K max output
  • no image, file input
Gemini 3.5 Flash Google1477.7 +56.5$1.49 summarise
  • Summarise −17%
  • Agentic −24%
  • 210K → 66K max output
Kimi K3 Moonshot AI1475.9 +54.7$1.23 summarise
  • Summarise −32%
  • Agentic under 1% less
  • no file input
Claude Sonnet 5.5 Anthropic1466.6 +45.4$2.13 agentic
  • Agentic −11%
  • 210K → 128K max output
Gemini 2.5 Pro Google1457.8 +36.6$1.37 summarise
  • Summarise −24%
  • Agentic −21%
  • 210K → 66K max output
GPT-6.1 Sol OpenAI1445.6 +24.4$2.07 agentic
  • Agentic −14%
  • 210K → 128K max output
Claude Sonnet 5 Anthropic1442.9 +21.7$2.13 agentic
  • Agentic −11%
  • 210K → 128K max output
GPT-5.1 OpenAI1422.2 tie$1.37 summarise
  • Summarise −24%
  • Agentic −21%
  • 210K → 128K max output

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