Enter opens the highlighted result. Esc closes.

DeepSeek V3.2 vs Mistral Medium 3.5

DeepSeek V3.2 leads Mistral Medium 3.5 on coding, 30.0 to 25.5 on the Vibe Coding Index, and DeepSeek V3.2 costs 8.5x less per output token ($0.88 vs $7.50 per million). Updated October 21, 2026; 56 benchmarks have a result for both.

Replace DeepSeek V3.2 Replace Mistral Medium 3.5
DeepSeek V3.2DeepSeek
Coding
30.0
Overall
37.9
Output $/M
$0.88
41
Mistral Medium 3.5Mistral AI
Coding
25.5
Overall
37.3
Output $/M
$7.50
15
vs boards won of 56 shared

DeepSeek V3.2 or Mistral Medium 3.5: which to pick

Pick DeepSeek V3.2 for coding: it scores 30.0 on the Vibe Coding Index against 25.5 for Mistral Medium 3.5.

Pick Mistral Medium 3.5 for a larger context window (262K against 164K tokens) and work measured by AA Coding Index or LMArena Text: coding, where it leads.

Model ADeepSeek V3.2DeepSeek30.0/1003 independent benchmarks$0.99 per solved taskModel BMistral Medium 3.5Mistral AI25.5/1003 independent benchmarks$6.02 per solved task
Shared results
56
DeepSeek V3.2 only
19
Mistral Medium 3.5 only
25
Like-for-like categories
8 of 9

Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.

Which one for your work

Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.

At a glance

Coding index
DeepSeek V3.2 30.0 from 3 benchmarks, Mistral Medium 3.5 25.5 from 3 benchmarks.
Widest category
Reasoning, 11.1 points apart on average over 6 shared benchmarks.
Widest single gap
GPQA Diamond: DeepSeek V3.2 80.3% against Mistral Medium 3.5 34.9%, 45.4 points apart on a 0 to 100 scale.
Output price
DeepSeek V3.2 $0.88 against Mistral Medium 3.5 $7.50 per million output tokens, 8.5x apart.
Context window
DeepSeek V3.2 takes 164K tokens, Mistral Medium 3.5 262K.

Benchmark by benchmark

First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.

Boards won, where both have a result
CategoryDeepSeek V3.2BarsMistral Medium 3.5Ties
Coding63
Intelligence20
Agentic work01
Reasoning51
Math11
Knowledge53
Web design80
General146
All shared boards4115
CodingLike-for-likeSix of these boards feed the Vibe Coding Index63
Coding benchmarks for DeepSeek V3.2 and Mistral Medium 3.5
BenchmarkDeepSeek V3.2Mistral Medium 3.5Leader
SWE-bench Verified70.0%Not measuredOnly DeepSeek V3.2 measured
SWE-bench Multilingual59.0%Not measuredOnly DeepSeek V3.2 measured
LMArena WebDev13621264DeepSeek V3.2
AA Coding Index44.246.9Mistral Medium 3.5
LiveCodeBench (Vals AI)80.7%Not measuredOnly DeepSeek V3.2 measured
SWE-bench Verified (Vals AI)67.6%66.4%DeepSeek V3.2
Vibe Code Bench (Vals AI)5.1%2.9%DeepSeek V3.2
Code migration (Vals AI)Not measured5.1%Only Mistral Medium 3.5 measured
Terminal-Bench 2.1 (Vals AI)Not measured39.0%Only Mistral Medium 3.5 measured
LiveCodeBench (Artificial Analysis run)86.2%Not measuredOnly DeepSeek V3.2 measured
SciCodeNot measured40.2%Only Mistral Medium 3.5 measured
Terminal-Bench Hard (Artificial Analysis run)35.6%33.3%DeepSeek V3.2
Terminal-Bench39.6%Not measuredOnly DeepSeek V3.2 measured
ALE-BenchNot measured764.0Only Mistral Medium 3.5 measured
LMArena Text: coding14531460Mistral Medium 3.5
LMArena Text: software and IT14491458Mistral Medium 3.5
LMArena WebDev: React13501254DeepSeek V3.2
LMArena WebDev: HTML13711308DeepSeek V3.2
Neither model has a result on the other 21 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for DeepSeek V3.2 and Mistral Medium 3.5
BenchmarkDeepSeek V3.2Mistral Medium 3.5Leader
AA Intelligence Index21.514.2DeepSeek V3.2
Epoch Capabilities Index146.1141.4DeepSeek V3.2
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
Agentic work benchmarks for DeepSeek V3.2 and Mistral Medium 3.5
BenchmarkDeepSeek V3.2Mistral Medium 3.5Leader
AA Agentic IndexNot measured7.0Only Mistral Medium 3.5 measured
Finance Agent (Vals AI)Not measured32.1%Only Mistral Medium 3.5 measured
Tax Agent Bench (Vals AI)Not measured4.5%Only Mistral Medium 3.5 measured
Legal Research Bench (Vals AI)Not measured9.1%Only Mistral Medium 3.5 measured
Excel Modeling Benchmark (Vals AI)Not measured11.2%Only Mistral Medium 3.5 measured
τ²-Bench Telecom90.6%94.2%Mistral Medium 3.5
APEX-Agents21.3%Not measuredOnly DeepSeek V3.2 measured
Vending-Bench 2$1,034Not measuredOnly DeepSeek V3.2 measured
Neither model has a result on the other 33 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index51
Reasoning benchmarks for DeepSeek V3.2 and Mistral Medium 3.5
BenchmarkDeepSeek V3.2Mistral Medium 3.5Leader
GPQA Diamond80.3%34.9%DeepSeek V3.2
GPQA Diamond (Artificial Analysis run)84.0%74.8%DeepSeek V3.2
Humanity's Last Exam (Artificial Analysis run)24.6%13.8%DeepSeek V3.2
ARC-AGI57.0%Not measuredOnly DeepSeek V3.2 measured
ARC-AGI-24.0%Not measuredOnly DeepSeek V3.2 measured
LMArena Text: hard prompts14341436Mistral Medium 3.5
LMArena Text: expert14361432DeepSeek V3.2
LMArena Text: hard prompts in English14461445DeepSeek V3.2
LMArena Vision: diagramsNot measured1254Only Mistral Medium 3.5 measured
LMArena Vision: homeworkNot measured1230Only Mistral Medium 3.5 measured
Kagi LLM BenchmarkNot measured41.4%Only Mistral Medium 3.5 measured
Neither model has a result on the other 11 boards here
MathLike-for-likePublic boards, shown as published, kept out of the index11
Math benchmarks for DeepSeek V3.2 and Mistral Medium 3.5
BenchmarkDeepSeek V3.2Mistral Medium 3.5Leader
AIME84.6%Not measuredOnly DeepSeek V3.2 measured
MGSM (Vals AI)86.0%Not measuredOnly DeepSeek V3.2 measured
AA Math Index92.0Not measuredOnly DeepSeek V3.2 measured
AIME 2025 (Artificial Analysis run)92.0%Not measuredOnly DeepSeek V3.2 measured
ProofBenchNot measured9.0%Only Mistral Medium 3.5 measured
LMArena Text: math14351431DeepSeek V3.2
LMArena Text: mathematical work14401440Mistral Medium 3.5
Neither model has a result on the other 6 boards here
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index53
Knowledge benchmarks for DeepSeek V3.2 and Mistral Medium 3.5
BenchmarkDeepSeek V3.2Mistral Medium 3.5Leader
MMLU Pro84.9%75.3%DeepSeek V3.2
LegalBench (Vals AI)76.1%Not measuredOnly DeepSeek V3.2 measured
CaseLaw (Vals AI)55.4%44.2%DeepSeek V3.2
CorpFin (Vals AI)51.0%58.8%Mistral Medium 3.5
TaxEval (Vals AI)68.2%68.0%DeepSeek V3.2
MortgageTax (Vals AI)Not measured28.9%Only Mistral Medium 3.5 measured
MedQA (Vals AI)93.9%Not measuredOnly DeepSeek V3.2 measured
MedCode (Vals AI)Not measured33.8%Only Mistral Medium 3.5 measured
MMLU Pro (Artificial Analysis run)86.2%Not measuredOnly DeepSeek V3.2 measured
LMArena Text: science14461428DeepSeek V3.2
LMArena Text: medicine and healthcare14401409DeepSeek V3.2
LMArena Text: legal and government14251437Mistral Medium 3.5
LMArena Text: business and finance14101426Mistral Medium 3.5
Neither model has a result on the other 5 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index80
Web design benchmarks for DeepSeek V3.2 and Mistral Medium 3.5
BenchmarkDeepSeek V3.2Mistral Medium 3.5Leader
Design Arena: websites11751142DeepSeek V3.2
Design Arena: UI components11491079DeepSeek V3.2
Design Arena: game dev11321079DeepSeek V3.2
Design Arena: data viz11581120DeepSeek V3.2
Design Arena: 3D11331061DeepSeek V3.2
Design Arena: all code categories11641121DeepSeek V3.2
Design Arena: SVG1032950DeepSeek V3.2
Design Arena: ASCII art10831063DeepSeek V3.2
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index146
General benchmarks for DeepSeek V3.2 and Mistral Medium 3.5
BenchmarkDeepSeek V3.2Mistral Medium 3.5Leader
SAGE (Vals AI)Not measured37.6%Only Mistral Medium 3.5 measured
MedScribe (Vals AI)Not measured67.7%Only Mistral Medium 3.5 measured
Vals Multimodal IndexNot measured34.8%Only Mistral Medium 3.5 measured
IFBench60.7%68.8%Mistral Medium 3.5
Long context reasoning (AA-LCR)73.3%69.3%DeepSeek V3.2
DTBench85.6%75.5%DeepSeek V3.2
LMCA28.8%26.1%DeepSeek V3.2
CL-bench12.4%Not measuredOnly DeepSeek V3.2 measured
Surface Evolver BenchNot measured26.9%Only Mistral Medium 3.5 measured
CL-bench Life7.4%Not measuredOnly DeepSeek V3.2 measured
LMArena Text14251421DeepSeek V3.2
LMArena Text: instruction following14131415Mistral Medium 3.5
LMArena Text: creative writing14001374DeepSeek V3.2
LMArena VisionNot measured1222Only Mistral Medium 3.5 measured
LMArena Text: English14391433DeepSeek V3.2
LMArena Text: non-English14091404DeepSeek V3.2
LMArena Text: Chinese14551444DeepSeek V3.2
LMArena Text: German14131453Mistral Medium 3.5
LMArena Text: Japanese1374Not measuredOnly DeepSeek V3.2 measured
LMArena Text: Korean13701384Mistral Medium 3.5
LMArena Text: French14291448Mistral Medium 3.5
LMArena Text: Spanish14241411DeepSeek V3.2
LMArena Text: Russian14241395DeepSeek V3.2
LMArena Text: Polish14191430Mistral Medium 3.5
LMArena Text: multi-turn14271423DeepSeek V3.2
LMArena Text: longer queries14281415DeepSeek V3.2
LMArena Text: writing and language14071384DeepSeek V3.2
LMArena Text: entertainment and media13951382DeepSeek V3.2
LMArena Vision: EnglishNot measured1240Only Mistral Medium 3.5 measured
LMArena Vision: ChineseNot measured1244Only Mistral Medium 3.5 measured
LMArena Vision: reading textNot measured1233Only Mistral Medium 3.5 measured
LMArena Vision: humorNot measured1216Only Mistral Medium 3.5 measured
LMArena Vision: creative writingNot measured1198Only Mistral Medium 3.5 measured
Neither model has a result on the other 13 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for DeepSeek V3.2 and Mistral Medium 3.5
BenchmarkDeepSeek V3.2Mistral Medium 3.5Leader
Neither model has a result on 5 boards here

What each workload costs

Three fixed token mixes at each model's listed price, so per-token rates become a bill.

Chat turn

1K input + 500 output tokens

DeepSeek V3.2$0.0007Fits in one request
Mistral Medium 3.5$0.0053Fits in one request

DeepSeek V3.2 costs less

Repository review

50K input + 3K output tokens

DeepSeek V3.2$0.017Fits in one request
Mistral Medium 3.5$0.098Fits in one request

DeepSeek V3.2 costs less

Agent loop

200K input + 10K output tokens

DeepSeek V3.2$0.067Exceeds the context window
Mistral Medium 3.5$0.375Fits in one request

DeepSeek V3.2 costs less

Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.

Specification differences

The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.

Specifications of DeepSeek V3.2 and Mistral Medium 3.5
SpecificationDeepSeek V3.2Mistral Medium 3.5
MakerDeepSeekMistral AI
WeightsOpen weightsNo public weights listed
ReleasedDecember 1, 2025April 30, 2026
Context window164K262K
Input $/Mtok$0.29$1.50
Output $/Mtok$0.88$7.50
Coding index30.025.5
Independent measurements behind it33
Cost per solved task$0.99$6.02
SourceOpenRouter listingOpenRouter listing

Questions

Is DeepSeek V3.2 better than Mistral Medium 3.5 for coding?

Pick DeepSeek V3.2 for coding: it scores 30.0 on the Vibe Coding Index against 25.5 for Mistral Medium 3.5. Pick Mistral Medium 3.5 for a larger context window (262K against 164K tokens) and work measured by AA Coding Index or LMArena Text: coding, where it leads.

Which is cheaper, DeepSeek V3.2 or Mistral Medium 3.5?

DeepSeek V3.2 costs $0.29 per million input tokens and $0.88 per million output tokens; Mistral Medium 3.5 costs $1.50 and $7.50. Per solved SWE-bench Verified task, DeepSeek V3.2 costs $0.99 and Mistral Medium 3.5 $6.02.

Which has the bigger context window, DeepSeek V3.2 or Mistral Medium 3.5?

Mistral Medium 3.5, with 262K tokens against 164K.

How many benchmarks is the DeepSeek V3.2 vs Mistral Medium 3.5 comparison based on?

56 benchmarks have a result for both models as of October 21, 2026. DeepSeek V3.2's coding index rests on 3 independent benchmarks and Mistral Medium 3.5's on 3. Where only one model has a result, the page says so instead of naming a winner.

Compare next

Embed this comparison on your site

Free under CC BY 4.0, while the source link stays visible. Preview the frame · Terms and other embeds

<iframe src="https://llmbenchmark.ai/embed/compare/deepseek-v3-2-vs-mistral-medium-3-5" title="DeepSeek V3.2 vs Mistral Medium 3.5, LLMBenchmark.ai" width="100%" height="360" style="border:0;max-width:100%" loading="lazy"></iframe>
<p style="font:13px/1.4 sans-serif;margin:6px 0 0">Source: <a href="https://llmbenchmark.ai/compare/deepseek-v3-2-vs-mistral-medium-3-5">LLMBenchmark.ai, DeepSeek V3.2 vs Mistral Medium 3.5</a></p>