Mistral Small 3.1 24B vs Hy3
Hy3 costs 1.7x less per output token than Mistral Small 3.1 24B ($0.33 vs $0.555 per million). Updated October 21, 2026; 28 benchmarks have a result for both.
Mistral Small 3.1 24B or Hy3: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Mistral Small 3.1 24B for work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.
- Shared results
- 28
- Mistral Small 3.1 24B only
- 10
- Hy3 only
- 16
- Like-for-like categories
- 5 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.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Web design
Design Arena: websites
No call
Only Hy3 has a Design Arena: websites result, 1230.
Directional only
Long documents
Listed context window
Hy3
Hy3 takes 262K tokens against 128K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Hy3
Mistral Small 3.1 24B $0.0006, Hy3 $0.0002 per turn.
List prices
At a glance
- Coding index
- Mistral Small 3.1 24B not benchmarked, Hy3 not benchmarked.
- Widest category
- Math, 35.2 points apart on average over 2 shared benchmarks.
- Widest single gap
- LMArena Text: Chinese: Mistral Small 3.1 24B 1253 against Hy3 1485, 38.7 points apart on a 0 to 100 scale.
- Output price
- Mistral Small 3.1 24B $0.555 against Hy3 $0.33 per million output tokens, 1.7x apart.
- Context window
- Mistral Small 3.1 24B takes 128K tokens, Hy3 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.
| Category | Mistral Small 3.1 24B | Bars | Hy3 | Ties |
|---|---|---|---|---|
| Coding | 0 | 2 | ||
| Reasoning | 0 | 3 | ||
| Math | 0 | 2 | ||
| Knowledge | 0 | 4 | ||
| General | 0 | 17 | ||
| All shared boards | 0 | 28 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index02
| Benchmark | Mistral Small 3.1 24B | Hy3 | Leader |
|---|---|---|---|
| LMArena WebDev | Not measured | 1507 | Only Hy3 measured |
| SciCode | Not measured | 48.6% | Only Hy3 measured |
| ALE-Bench | Not measured | 838.6 | Only Hy3 measured |
| LMArena Text: coding | 1309 | 1466 | Hy3 |
| LMArena Text: software and IT | 1301 | 1463 | Hy3 |
| LMArena WebDev: React | Not measured | 1511 | Only Hy3 measured |
| LMArena WebDev: HTML | Not measured | 1487 | Only Hy3 measured |
| Neither model has a result on the other 32 boards here | |||
IntelligenceNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Mistral Small 3.1 24B | Hy3 | Leader |
|---|---|---|---|
| Neither model has a result on 2 boards here | |||
Agentic workNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Mistral Small 3.1 24B | Hy3 | Leader |
|---|---|---|---|
| Neither model has a result on 41 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index03
| Benchmark | Mistral Small 3.1 24B | Hy3 | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | Not measured | 89.7% | Only Hy3 measured |
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 33.5% | Only Hy3 measured |
| LMArena Text: hard prompts | 1278 | 1447 | Hy3 |
| LMArena Text: expert | 1257 | 1461 | Hy3 |
| LMArena Text: hard prompts in English | 1290 | 1455 | Hy3 |
| LMArena Vision: diagrams | 1119 | Not measured | Only Mistral Small 3.1 24B measured |
| LMArena Vision: homework | 1158 | Not measured | Only Mistral Small 3.1 24B measured |
| Neither model has a result on the other 15 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Mistral Small 3.1 24B | Hy3 | Leader |
|---|---|---|---|
| LMArena Text: math | 1261 | 1473 | Hy3 |
| LMArena Text: mathematical work | 1271 | 1482 | Hy3 |
| Neither model has a result on the other 11 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
| Benchmark | Mistral Small 3.1 24B | Hy3 | Leader |
|---|---|---|---|
| LMArena Text: science | 1265 | 1457 | Hy3 |
| LMArena Text: medicine and healthcare | 1267 | 1452 | Hy3 |
| LMArena Text: legal and government | 1271 | 1422 | Hy3 |
| LMArena Text: business and finance | 1257 | 1424 | Hy3 |
| Neither model has a result on the other 14 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | Mistral Small 3.1 24B | Hy3 | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1230 | Only Hy3 measured |
| Design Arena: UI components | Not measured | 1207 | Only Hy3 measured |
| Design Arena: game dev | Not measured | 1208 | Only Hy3 measured |
| Design Arena: data viz | Not measured | 1163 | Only Hy3 measured |
| Design Arena: 3D | Not measured | 1233 | Only Hy3 measured |
| Design Arena: all code categories | Not measured | 1227 | Only Hy3 measured |
| Design Arena: SVG | Not measured | 1200 | Only Hy3 measured |
| Design Arena: ASCII art | Not measured | 1152 | Only Hy3 measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index017
| Benchmark | Mistral Small 3.1 24B | Hy3 | Leader |
|---|---|---|---|
| Long context reasoning (AA-LCR) | Not measured | 79.0% | Only Hy3 measured |
| LMArena Text | 1277 | 1438 | Hy3 |
| LMArena Text: instruction following | 1263 | 1426 | Hy3 |
| LMArena Text: creative writing | 1252 | 1403 | Hy3 |
| LMArena Vision | 1136 | Not measured | Only Mistral Small 3.1 24B measured |
| LMArena Text: English | 1295 | 1448 | Hy3 |
| LMArena Text: non-English | 1255 | 1426 | Hy3 |
| LMArena Text: Chinese | 1253 | 1485 | Hy3 |
| LMArena Text: German | 1266 | 1449 | Hy3 |
| LMArena Text: Japanese | 1208 | 1397 | Hy3 |
| LMArena Text: Korean | 1207 | 1396 | Hy3 |
| LMArena Text: French | 1273 | 1456 | Hy3 |
| LMArena Text: Spanish | 1284 | 1457 | Hy3 |
| LMArena Text: Russian | 1263 | 1432 | Hy3 |
| LMArena Text: Polish | 1242 | 1447 | Hy3 |
| LMArena Text: multi-turn | 1270 | 1436 | Hy3 |
| LMArena Text: longer queries | 1299 | 1442 | Hy3 |
| LMArena Text: writing and language | 1268 | 1414 | Hy3 |
| LMArena Text: entertainment and media | 1250 | 1400 | Hy3 |
| LMArena Vision: English | 1157 | Not measured | Only Mistral Small 3.1 24B measured |
| LMArena Vision: Chinese | 1155 | Not measured | Only Mistral Small 3.1 24B measured |
| LMArena Vision: reading text | 1132 | Not measured | Only Mistral Small 3.1 24B measured |
| LMArena Vision: humor | 1142 | Not measured | Only Mistral Small 3.1 24B measured |
| LMArena Vision: captioning | 1153 | Not measured | Only Mistral Small 3.1 24B measured |
| LMArena Vision: recognizing things | 1155 | Not measured | Only Mistral Small 3.1 24B measured |
| LMArena Vision: creative writing | 1123 | Not measured | Only Mistral Small 3.1 24B measured |
| Neither model has a result on the other 20 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Mistral Small 3.1 24B | Hy3 | Leader |
|---|---|---|---|
| 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
Hy3 costs less
Repository review
50K input + 3K output tokens
Hy3 costs less
Agent loop
200K input + 10K output tokens
Hy3 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.
| Specification | Mistral Small 3.1 24B | Tencent Hy3 |
|---|---|---|
| Maker | Mistral AI | Tencent |
| Weights | Open weights | Open weights |
| Released | March 17, 2025 | July 6, 2026 |
| Context window | 128K | 262K |
| Input $/Mtok | $0.351 | $0.0825 |
| Output $/Mtok | $0.555 | $0.33 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 1 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Mistral Small 3.1 24B better than Hy3 for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Mistral Small 3.1 24B for work measured by LMArena Text: coding or LMArena Text: software and IT, where it leads.
Which is cheaper, Mistral Small 3.1 24B or Hy3?
Mistral Small 3.1 24B costs $0.351 per million input tokens and $0.555 per million output tokens; Hy3 costs $0.0825 and $0.33. Cost per solved task is known for neither of them.
Which has the bigger context window, Mistral Small 3.1 24B or Hy3?
Hy3, with 262K tokens against 128K.
How many benchmarks is the Mistral Small 3.1 24B vs Hy3 comparison based on?
28 benchmarks have a result for both models as of October 21, 2026. Mistral Small 3.1 24B's coding index rests on 0 independent benchmarks and Hy3's on 1. Where only one model has a result, the page says so instead of naming a winner.