Enter opens the highlighted result. Esc closes.

GPT-5.3-Codex vs Mistral Medium 3

GPT-5.3-Codex leads Mistral Medium 3 on the Artificial Analysis Intelligence Index, 32.5 to 9.0, and Mistral Medium 3 costs 7.0x less per output token ($2.00 vs $14.00 per million). Updated October 21, 2026; 14 benchmarks have a result for both.

Replace GPT-5.3-Codex Replace Mistral Medium 3
GPT-5.3-CodexOpenAI
Coding
60.0
Overall
No data
Output $/M
$14.00
13
Mistral Medium 3Mistral AI
Coding
No data
Overall
No data
Output $/M
$2.00
1
vs boards won of 14 shared

GPT-5.3-Codex or Mistral Medium 3: which to pick

Only GPT-5.3-Codex has a coding score, 60.0 on the Vibe Coding Index; Mistral Medium 3 has none, so the two cannot be ranked on coding.

Pick Mistral Medium 3 for output tokens 7.0x cheaper ($2.00 vs $14.00 per million) and work measured by Design Arena: 3D, where it leads.

Model AGPT-5.3-CodexOpenAI60.0/1003 independent benchmarks$0.59 per solved taskModel BMistral Medium 3Mistral AINot benchmarkedNo coding benchmarkNo cost per solved task
Shared results
14
GPT-5.3-Codex only
20
Mistral Medium 3 only
43
Like-for-like categories
6 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
GPT-5.3-Codex 60.0 from 3 benchmarks, Mistral Medium 3 not benchmarked.
Widest category
Agentic work, 61.7 points apart on average over 1 shared benchmark.
Widest single gap
τ²-Bench Telecom: GPT-5.3-Codex 86.0% against Mistral Medium 3 24.3%, 61.7 points apart on a 0 to 100 scale.
Output price
GPT-5.3-Codex $14.00 against Mistral Medium 3 $2.00 per million output tokens, 7.0x apart.
Context window
GPT-5.3-Codex takes 400K tokens, Mistral Medium 3 131K.

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
CategoryGPT-5.3-CodexBarsMistral Medium 3Ties
Coding10
Intelligence20
Agentic work10
Reasoning20
Web design51
General20
All shared boards131
CodingLike-for-likeSix of these boards feed the Vibe Coding Index10
Coding benchmarks for GPT-5.3-Codex and Mistral Medium 3
BenchmarkGPT-5.3-CodexMistral Medium 3Leader
LMArena WebDev1407Not measuredOnly GPT-5.3-Codex measured
LiveCodeBench (Vals AI)87.3%Not measuredOnly GPT-5.3-Codex measured
SWE-bench Verified (Vals AI)78.0%Not measuredOnly GPT-5.3-Codex measured
Vibe Code Bench (Vals AI)61.8%Not measuredOnly GPT-5.3-Codex measured
IOI (Vals AI)53.8%Not measuredOnly GPT-5.3-Codex measured
LiveCodeBench (Artificial Analysis run)Not measured40.0%Only Mistral Medium 3 measured
Terminal-Bench Hard (Artificial Analysis run)53.0%3.8%GPT-5.3-Codex
Terminal-Bench78.4%Not measuredOnly GPT-5.3-Codex measured
WeirdML79.3%Not measuredOnly GPT-5.3-Codex measured
ALE-Bench1655.2Not measuredOnly GPT-5.3-Codex measured
LMArena Text: codingNot measured1385Only Mistral Medium 3 measured
LMArena Text: software and ITNot measured1385Only Mistral Medium 3 measured
LMArena WebDev: React1395Not measuredOnly GPT-5.3-Codex measured
LMArena WebDev: HTML1422Not measuredOnly GPT-5.3-Codex measured
LMArena WebDev: from an image1443Not measuredOnly GPT-5.3-Codex measured
Neither model has a result on the other 24 boards here
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
Intelligence benchmarks for GPT-5.3-Codex and Mistral Medium 3
BenchmarkGPT-5.3-CodexMistral Medium 3Leader
AA Intelligence Index32.59.0GPT-5.3-Codex
Epoch Capabilities Index157.1134.0GPT-5.3-Codex
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
Agentic work benchmarks for GPT-5.3-Codex and Mistral Medium 3
BenchmarkGPT-5.3-CodexMistral Medium 3Leader
Design Arena: agents building web apps1032Not measuredOnly GPT-5.3-Codex measured
Design Arena: agents building full-stack apps983Not measuredOnly GPT-5.3-Codex measured
Design Arena: agents building mobile apps1054Not measuredOnly GPT-5.3-Codex measured
Design Arena: agents building Android apps1084Not measuredOnly GPT-5.3-Codex measured
Design Arena: agents building Godot games1124Not measuredOnly GPT-5.3-Codex measured
τ²-Bench Telecom86.0%24.3%GPT-5.3-Codex
METR time horizons74.5%Not measuredOnly GPT-5.3-Codex measured
Vending-Bench 2$5,940Not measuredOnly GPT-5.3-Codex measured
Neither model has a result on the other 33 boards here
ReasoningLike-for-likePublic boards, shown as published, kept out of the index20
Reasoning benchmarks for GPT-5.3-Codex and Mistral Medium 3
BenchmarkGPT-5.3-CodexMistral Medium 3Leader
GPQA Diamond (Artificial Analysis run)91.5%57.8%GPT-5.3-Codex
Humanity's Last Exam (Artificial Analysis run)42.5%4.1%GPT-5.3-Codex
LMArena Text: hard promptsNot measured1364Only Mistral Medium 3 measured
LMArena Text: expertNot measured1342Only Mistral Medium 3 measured
LMArena Text: hard prompts in EnglishNot measured1374Only Mistral Medium 3 measured
LMArena Vision: diagramsNot measured1155Only Mistral Medium 3 measured
LMArena Vision: homeworkNot measured1165Only Mistral Medium 3 measured
Neither model has a result on the other 15 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index4 boards, none shared
Math benchmarks for GPT-5.3-Codex and Mistral Medium 3
BenchmarkGPT-5.3-CodexMistral Medium 3Leader
AA Math IndexNot measured30.3Only Mistral Medium 3 measured
AIME 2025 (Artificial Analysis run)Not measured30.3%Only Mistral Medium 3 measured
LMArena Text: mathNot measured1353Only Mistral Medium 3 measured
LMArena Text: mathematical workNot measured1354Only Mistral Medium 3 measured
Neither model has a result on the other 9 boards here
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index5 boards, none shared
Knowledge benchmarks for GPT-5.3-Codex and Mistral Medium 3
BenchmarkGPT-5.3-CodexMistral Medium 3Leader
MMLU Pro (Artificial Analysis run)Not measured76.0%Only Mistral Medium 3 measured
LMArena Text: scienceNot measured1381Only Mistral Medium 3 measured
LMArena Text: medicine and healthcareNot measured1381Only Mistral Medium 3 measured
LMArena Text: legal and governmentNot measured1384Only Mistral Medium 3 measured
LMArena Text: business and financeNot measured1352Only Mistral Medium 3 measured
Neither model has a result on the other 13 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index51
Web design benchmarks for GPT-5.3-Codex and Mistral Medium 3
BenchmarkGPT-5.3-CodexMistral Medium 3Leader
Design Arena: websites11631079GPT-5.3-Codex
Design Arena: UI components11371024GPT-5.3-Codex
Design Arena: game dev11611018GPT-5.3-Codex
Design Arena: data viz11641034GPT-5.3-Codex
Design Arena: 3D10081084Mistral Medium 3
Design Arena: all code categories11451069GPT-5.3-Codex
Design Arena: SVG1129Not measuredOnly GPT-5.3-Codex measured
Design Arena: ASCII art1152Not measuredOnly GPT-5.3-Codex measured
Neither model has a result on the other 1 board here
GeneralLike-for-likePublic boards, shown as published, kept out of the index20
General benchmarks for GPT-5.3-Codex and Mistral Medium 3
BenchmarkGPT-5.3-CodexMistral Medium 3Leader
IFBench75.4%39.3%GPT-5.3-Codex
Long context reasoning (AA-LCR)83.3%31.3%GPT-5.3-Codex
DTBenchNot measured62.3%Only Mistral Medium 3 measured
LMArena TextNot measured1370Only Mistral Medium 3 measured
LMArena Text: instruction followingNot measured1338Only Mistral Medium 3 measured
LMArena Text: creative writingNot measured1343Only Mistral Medium 3 measured
LMArena VisionNot measured1157Only Mistral Medium 3 measured
LMArena Text: EnglishNot measured1385Only Mistral Medium 3 measured
LMArena Text: non-EnglishNot measured1353Only Mistral Medium 3 measured
LMArena Text: ChineseNot measured1374Only Mistral Medium 3 measured
LMArena Text: GermanNot measured1377Only Mistral Medium 3 measured
LMArena Text: JapaneseNot measured1316Only Mistral Medium 3 measured
LMArena Text: KoreanNot measured1304Only Mistral Medium 3 measured
LMArena Text: FrenchNot measured1368Only Mistral Medium 3 measured
LMArena Text: SpanishNot measured1364Only Mistral Medium 3 measured
LMArena Text: RussianNot measured1359Only Mistral Medium 3 measured
LMArena Text: PolishNot measured1371Only Mistral Medium 3 measured
LMArena Text: multi-turnNot measured1384Only Mistral Medium 3 measured
LMArena Text: longer queriesNot measured1360Only Mistral Medium 3 measured
LMArena Text: writing and languageNot measured1352Only Mistral Medium 3 measured
LMArena Text: entertainment and mediaNot measured1335Only Mistral Medium 3 measured
LMArena Vision: EnglishNot measured1172Only Mistral Medium 3 measured
LMArena Vision: ChineseNot measured1161Only Mistral Medium 3 measured
LMArena Vision: reading textNot measured1154Only Mistral Medium 3 measured
LMArena Vision: humorNot measured1145Only Mistral Medium 3 measured
LMArena Vision: captioningNot measured1094Only Mistral Medium 3 measured
LMArena Vision: recognizing thingsNot measured1141Only Mistral Medium 3 measured
LMArena Vision: creative writingNot measured1146Only Mistral Medium 3 measured
Neither model has a result on the other 18 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for GPT-5.3-Codex and Mistral Medium 3
BenchmarkGPT-5.3-CodexMistral Medium 3Leader
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

GPT-5.3-Codex$0.0088Fits in one request
Mistral Medium 3$0.0014Fits in one request

Mistral Medium 3 costs less

Repository review

50K input + 3K output tokens

GPT-5.3-Codex$0.130Fits in one request
Mistral Medium 3$0.026Fits in one request

Mistral Medium 3 costs less

Agent loop

200K input + 10K output tokens

GPT-5.3-Codex$0.490Fits in one request
Mistral Medium 3$0.100Exceeds the context window

Mistral Medium 3 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 GPT-5.3-Codex and Mistral Medium 3
SpecificationOpenAI GPT-5.3-CodexMistral Medium 3
MakerOpenAIMistral AI
WeightsNo public weights listedNo public weights listed
ReleasedFebruary 24, 2026May 7, 2025
Context window400K131K
Input $/Mtok$1.75$0.40
Output $/Mtok$14.00$2.00
Coding index60.0Not benchmarked
Independent measurements behind it30
Cost per solved task$0.59Not computable
SourceOpenRouter listingOpenRouter listing

Questions

Is GPT-5.3-Codex better than Mistral Medium 3 for coding?

Only GPT-5.3-Codex has a coding score, 60.0 on the Vibe Coding Index; Mistral Medium 3 has none, so the two cannot be ranked on coding. Pick Mistral Medium 3 for output tokens 7.0x cheaper ($2.00 vs $14.00 per million) and work measured by Design Arena: 3D, where it leads.

Which is cheaper, GPT-5.3-Codex or Mistral Medium 3?

GPT-5.3-Codex costs $1.75 per million input tokens and $14.00 per million output tokens; Mistral Medium 3 costs $0.40 and $2.00. Cost per solved task is known for GPT-5.3-Codex only.

Which has the bigger context window, GPT-5.3-Codex or Mistral Medium 3?

GPT-5.3-Codex, with 400K tokens against 131K.

How many benchmarks is the GPT-5.3-Codex vs Mistral Medium 3 comparison based on?

14 benchmarks have a result for both models as of October 21, 2026. GPT-5.3-Codex's coding index rests on 3 independent benchmarks and Mistral Medium 3's on 0. Where only one model has a result, the page says so instead of naming a winner.

Compare next