Enter opens the highlighted result. Esc closes.

Mistral Large 4 vs Hy4 flash

Mistral Large 4 leads Hy4 flash on coding, 61.1 to 44.5 on the Vibe Coding Index, and Hy4 flash costs 14.9x less per output token ($0.1405 vs $2.10 per million). Updated November 22, 2026; 12 benchmarks have a result for both.

Mistral Large 4Mistral AI
Coding
61.1
Overall
57.9
Output $/M
$2.10
10
Hy4 flashTencent
Coding
44.5
Overall
No data
Output $/M
$0.1405
2
vs boards won of 12 shared

Mistral Large 4 or Hy4 flash: which to pick

Pick Mistral Large 4 for coding: it scores 61.1 on the Vibe Coding Index against 44.5 for Hy4 flash.

Pick Hy4 flash for output tokens 14.9x cheaper ($0.1405 vs $2.10 per million) and work measured by Legal Research Bench (Vals AI) or Design Arena: game dev, where it leads.

Shared results
12
Mistral Large 4 only
79
Hy4 flash only
0
Like-for-like categories
4 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
Mistral Large 4 61.1 from 5 benchmarks, Hy4 flash 44.5 from 2 benchmarks.
Widest category
Coding, 10.3 points apart on average over 3 shared benchmarks.
Widest single gap
Terminal-Bench 4 (Vals AI): Mistral Large 4 27.8% against Hy4 flash 15.2%, 16.8 points apart on a 0 to 100 scale.
Output price
Mistral Large 4 $2.10 against Hy4 flash $0.1405 per million output tokens, 14.9x apart.
Context window
Both take 1.05M tokens.

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
CategoryMistral Large 4BarsHy4 flashTies
Coding30
Agentic work21
Web design41
General10
All shared boards102
CodingLike-for-likeSix of these boards feed the Vibe Coding Index30
Coding benchmarks for Mistral Large 4 and Hy4 flash
BenchmarkMistral Large 4Hy4 flashLeader
LiveBench coding79.4%Not measuredOnly Mistral Large 4 measured
LMArena WebDev1573Not measuredOnly Mistral Large 4 measured
Terminal-Bench 4 (Vals AI)27.8%15.2%Mistral Large 4
Vibe Code Bench (Vals AI)81.6%74.8%Mistral Large 4
Code migration (Vals AI)30.6%Not measuredOnly Mistral Large 4 measured
IOI (Vals AI)45.3%Not measuredOnly Mistral Large 4 measured
SWE-bench Multimodal (Vals AI)33.1%Not measuredOnly Mistral Large 4 measured
Vibe Code Bench 1-100 (Vals AI)14.3%Not measuredOnly Mistral Large 4 measured
Terminal-Bench 4.0 (Artificial Analysis run)24.7%Not measuredOnly Mistral Large 4 measured
FrontierCode 2.015.0%Not measuredOnly Mistral Large 4 measured
FrontierCode39.4%Not measuredOnly Mistral Large 4 measured
MirrorCode40.0%Not measuredOnly Mistral Large 4 measured
ALE-Bench1213.2Not measuredOnly Mistral Large 4 measured
LMArena Text: coding1459Not measuredOnly Mistral Large 4 measured
LMArena Text: software and IT1459Not measuredOnly Mistral Large 4 measured
LMArena WebDev: React1582Not measuredOnly Mistral Large 4 measured
LMArena WebDev: HTML1560Not measuredOnly Mistral Large 4 measured
GameBench 257.950.5Mistral Large 4
Terminal-Bench 4.027.9%Not measuredOnly Mistral Large 4 measured
Neither model has a result on the other 33 boards here
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
Intelligence benchmarks for Mistral Large 4 and Hy4 flash
BenchmarkMistral Large 4Hy4 flashLeader
AA Intelligence Index26.1Not measuredOnly Mistral Large 4 measured
Epoch Capabilities Index154.6Not measuredOnly Mistral Large 4 measured
Agentic workLike-for-likePublic boards, shown as published, kept out of the index21
Agentic work benchmarks for Mistral Large 4 and Hy4 flash
BenchmarkMistral Large 4Hy4 flashLeader
LiveBench research80.3%Not measuredOnly Mistral Large 4 measured
LiveBench browser17.4%Not measuredOnly Mistral Large 4 measured
LiveBench long agentic24.7%Not measuredOnly Mistral Large 4 measured
Finance Agent (Vals AI)54.7%Not measuredOnly Mistral Large 4 measured
Tax Agent Bench (Vals AI)30.7%23.3%Mistral Large 4
Legal Research Bench (Vals AI)35.6%38.5%Hy4 flash
Harvey's Legal Agent Benchmark (Vals AI)17.5%5.4%Mistral Large 4
Excel Modeling Benchmark (Vals AI)56.2%Not measuredOnly Mistral Large 4 measured
BioMysteryBench (Vals AI)67.0%Not measuredOnly Mistral Large 4 measured
APEX-Agents51.6%Not measuredOnly Mistral Large 4 measured
Neither model has a result on the other 34 boards here
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index12 boards, none shared
Reasoning benchmarks for Mistral Large 4 and Hy4 flash
BenchmarkMistral Large 4Hy4 flashLeader
LiveBench reasoning85.2%Not measuredOnly Mistral Large 4 measured
MysteryMechanism (Vals AI)12.2%Not measuredOnly Mistral Large 4 measured
HLE-Diamond (Artificial Analysis run)17.3%Not measuredOnly Mistral Large 4 measured
CritPt v2, hard subset (Artificial Analysis run)27.0%Not measuredOnly Mistral Large 4 measured
GPQA Diamond (Epoch AI run)88.1%Not measuredOnly Mistral Large 4 measured
Chess puzzles33.0%Not measuredOnly Mistral Large 4 measured
Mystery game puzzles17.0%Not measuredOnly Mistral Large 4 measured
LMArena Text: hard prompts1436Not measuredOnly Mistral Large 4 measured
LMArena Text: expert1450Not measuredOnly Mistral Large 4 measured
LMArena Text: hard prompts in English1442Not measuredOnly Mistral Large 4 measured
LMArena Vision: diagrams1273Not measuredOnly Mistral Large 4 measured
LMArena Vision: homework1266Not measuredOnly Mistral Large 4 measured
Neither model has a result on the other 14 boards here
MathDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
Math benchmarks for Mistral Large 4 and Hy4 flash
BenchmarkMistral Large 4Hy4 flashLeader
OTIS Mock AIME98.3%Not measuredOnly Mistral Large 4 measured
FrontierMath, tiers 1 to 358.6%Not measuredOnly Mistral Large 4 measured
FrontierMath, tier 426.8%Not measuredOnly Mistral Large 4 measured
ProofBench10.0%Not measuredOnly Mistral Large 4 measured
LMArena Text: math1445Not measuredOnly Mistral Large 4 measured
LMArena Text: mathematical work1456Not measuredOnly Mistral Large 4 measured
MathArena20.7%Not measuredOnly Mistral Large 4 measured
Neither model has a result on the other 7 boards here
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
Knowledge benchmarks for Mistral Large 4 and Hy4 flash
BenchmarkMistral Large 4Hy4 flashLeader
MedCode (Vals AI)40.7%Not measuredOnly Mistral Large 4 measured
AA-Omniscience5.1Not measuredOnly Mistral Large 4 measured
SimpleQA Verified20.0%Not measuredOnly Mistral Large 4 measured
LMArena Text: science1434Not measuredOnly Mistral Large 4 measured
LMArena Text: medicine and healthcare1403Not measuredOnly Mistral Large 4 measured
LMArena Text: legal and government1427Not measuredOnly Mistral Large 4 measured
LMArena Text: business and finance1430Not measuredOnly Mistral Large 4 measured
Neither model has a result on the other 14 boards here
Web designLike-for-likePublic boards, shown as published, kept out of the index41
Web design benchmarks for Mistral Large 4 and Hy4 flash
BenchmarkMistral Large 4Hy4 flashLeader
Design Arena: websites12531227Mistral Large 4
Design Arena: UI components12721262Mistral Large 4
Design Arena: game dev11901202Hy4 flash
Design Arena: data viz1271Not measuredOnly Mistral Large 4 measured
Design Arena: 3D12251209Mistral Large 4
Design Arena: all code categories12471225Mistral Large 4
Design Arena: SVG1216Not measuredOnly Mistral Large 4 measured
Design Arena: ASCII art1196Not measuredOnly Mistral Large 4 measured
LiveBench design76.3%Not measuredOnly Mistral Large 4 measured
GeneralLike-for-likePublic boards, shown as published, kept out of the index10
General benchmarks for Mistral Large 4 and Hy4 flash
BenchmarkMistral Large 4Hy4 flashLeader
Vals Index56.5%50.8%Mistral Large 4
LiveBench data analysis76.6%Not measuredOnly Mistral Large 4 measured
MedScribe (Vals AI)80.4%Not measuredOnly Mistral Large 4 measured
DTBench88.2%Not measuredOnly Mistral Large 4 measured
LMCA35.8%Not measuredOnly Mistral Large 4 measured
Surface Evolver Bench28.1%Not measuredOnly Mistral Large 4 measured
Furniture assembly20.8%Not measuredOnly Mistral Large 4 measured
LMArena Text1413Not measuredOnly Mistral Large 4 measured
LMArena Text: instruction following1414Not measuredOnly Mistral Large 4 measured
LMArena Text: creative writing1341Not measuredOnly Mistral Large 4 measured
LMArena Vision1246Not measuredOnly Mistral Large 4 measured
LMArena Text: English1423Not measuredOnly Mistral Large 4 measured
LMArena Text: non-English1400Not measuredOnly Mistral Large 4 measured
LMArena Text: Chinese1466Not measuredOnly Mistral Large 4 measured
LMArena Text: French1443Not measuredOnly Mistral Large 4 measured
LMArena Text: Spanish1456Not measuredOnly Mistral Large 4 measured
LMArena Text: Russian1415Not measuredOnly Mistral Large 4 measured
LMArena Text: multi-turn1423Not measuredOnly Mistral Large 4 measured
LMArena Text: longer queries1409Not measuredOnly Mistral Large 4 measured
LMArena Text: writing and language1367Not measuredOnly Mistral Large 4 measured
LMArena Text: entertainment and media1358Not measuredOnly Mistral Large 4 measured
LMArena Vision: English1258Not measuredOnly Mistral Large 4 measured
LMArena Vision: Chinese1331Not measuredOnly Mistral Large 4 measured
LMArena Vision: reading text1261Not measuredOnly Mistral Large 4 measured
LMArena Vision: creative writing1213Not measuredOnly Mistral Large 4 measured
Neither model has a result on the other 24 boards here
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
Image generation benchmarks for Mistral Large 4 and Hy4 flash
BenchmarkMistral Large 4Hy4 flashLeader
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

Mistral Large 4$0.0018Fits in one request
Hy4 flash$0.0001Fits in one request

Hy4 flash costs less

Repository review

50K input + 3K output tokens

Mistral Large 4$0.041Fits in one request
Hy4 flash$0.0022Fits in one request

Hy4 flash costs less

Agent loop

200K input + 10K output tokens

Mistral Large 4$0.161Fits in one request
Hy4 flash$0.0084Fits in one request

Hy4 flash 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 Mistral Large 4 and Hy4 flash
SpecificationMistral Large 4Tencent Hy4 flash
MakerMistral AITencent
WeightsOpen weightsOpen weights
ReleasedOctober 6, 2026November 16, 2026
Context window1.05M1.05M
Input $/Mtok$0.70$0.035
Output $/Mtok$2.10$0.1405
Coding index61.144.5
Independent measurements behind it52
Cost per solved taskNot computableNot computable
SourceOpenRouter listingOpenRouter listing

Questions

Is Mistral Large 4 better than Hy4 flash for coding?

Pick Mistral Large 4 for coding: it scores 61.1 on the Vibe Coding Index against 44.5 for Hy4 flash. Pick Hy4 flash for output tokens 14.9x cheaper ($0.1405 vs $2.10 per million) and work measured by Legal Research Bench (Vals AI) or Design Arena: game dev, where it leads.

Which is cheaper, Mistral Large 4 or Hy4 flash?

Mistral Large 4 costs $0.70 per million input tokens and $2.10 per million output tokens; Hy4 flash costs $0.035 and $0.1405. Cost per solved task is known for neither of them.

Which has the bigger context window, Mistral Large 4 or Hy4 flash?

Neither: both take 1.05M tokens.

How many benchmarks is the Mistral Large 4 vs Hy4 flash comparison based on?

12 benchmarks have a result for both models as of November 22, 2026. Mistral Large 4's coding index rests on 5 independent benchmarks and Hy4 flash's on 2. 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/mistral-large-4-0-vs-tencent-hy4-flash" title="Mistral Large 4 vs Hy4 flash, 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/mistral-large-4-0-vs-tencent-hy4-flash">LLMBenchmark.ai, Mistral Large 4 vs Hy4 flash</a></p>