GLM 4.7 Flash vs Hy3 preview
Hy3 preview leads GLM 4.7 Flash on the Artificial Analysis Intelligence Index, 25.3 to 14.9, and GLM 4.7 Flash costs 1.5x less per output token ($0.40 vs $0.60 per million). Updated October 21, 2026; 31 benchmarks have a result for both.
GLM 4.7 Flash or Hy3 preview: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick GLM 4.7 Flash for output tokens 1.5x cheaper ($0.40 vs $0.60 per million) and work measured by Terminal-Bench Hard (Artificial Analysis run) or LMArena Text: coding, where it leads.
- Shared results
- 31
- GLM 4.7 Flash only
- 12
- Hy3 preview only
- 5
- Like-for-like categories
- 7 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
Intelligence
AA Intelligence Index
Hy3 preview
Hy3 preview leads, 25.3 against 14.9.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Hy3 preview has a AA Agentic Index result, 24.1.
Directional only
Web design
Design Arena: websites
No call
Only GLM 4.7 Flash has a Design Arena: websites result, 1194.
Directional only
Long documents
Listed context window
Hy3 preview
Hy3 preview takes 262K tokens against 200K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
GLM 4.7 Flash
GLM 4.7 Flash $0.0003, Hy3 preview $0.0005 per turn.
List prices
At a glance
- Coding index
- GLM 4.7 Flash not benchmarked, Hy3 preview not benchmarked.
- Widest category
- Reasoning, 15.4 points apart on average over 5 shared benchmarks.
- Widest single gap
- GPQA Diamond (Artificial Analysis run): GLM 4.7 Flash 58.1% against Hy3 preview 86.7%, 28.6 points apart on a 0 to 100 scale.
- Output price
- GLM 4.7 Flash $0.40 against Hy3 preview $0.60 per million output tokens, 1.5x apart.
- Context window
- GLM 4.7 Flash takes 200K tokens, Hy3 preview 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 | GLM 4.7 Flash | Bars | Hy3 preview | Ties |
|---|---|---|---|---|
| Coding | 0 | 3 | ||
| Intelligence | 0 | 1 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 0 | 5 | ||
| Math | 0 | 2 | ||
| Knowledge | 0 | 4 | ||
| General | 0 | 15 | ||
| All shared boards | 1 | 30 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index03
| Benchmark | GLM 4.7 Flash | Hy3 preview | Leader |
|---|---|---|---|
| LMArena WebDev | Not measured | 1357 | Only Hy3 preview measured |
| AA Coding Index | Not measured | 58.8 | Only Hy3 preview measured |
| Terminal-Bench Hard (Artificial Analysis run) | 22.0% | 34.1% | Hy3 preview |
| LMArena Text: coding | 1382 | 1421 | Hy3 preview |
| LMArena Text: software and IT | 1389 | 1423 | Hy3 preview |
| LMArena WebDev: React | Not measured | 1349 | Only Hy3 preview measured |
| LMArena WebDev: HTML | Not measured | 1356 | Only Hy3 preview measured |
| Neither model has a result on the other 32 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | GLM 4.7 Flash | Hy3 preview | Leader |
|---|---|---|---|
| AA Intelligence Index | 14.9 | 25.3 | Hy3 preview |
| Neither model has a result on the other 1 board here | |||
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | GLM 4.7 Flash | Hy3 preview | Leader |
|---|---|---|---|
| AA Agentic Index | Not measured | 24.1 | Only Hy3 preview measured |
| τ²-Bench Telecom | 98.8% | 92.7% | GLM 4.7 Flash |
| Neither model has a result on the other 39 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index05
| Benchmark | GLM 4.7 Flash | Hy3 preview | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 58.1% | 86.7% | Hy3 preview |
| Humanity's Last Exam (Artificial Analysis run) | 7.6% | 27.8% | Hy3 preview |
| GPQA Diamond (Epoch AI run) | 60.5% | Not measured | Only GLM 4.7 Flash measured |
| LMArena Text: hard prompts | 1356 | 1413 | Hy3 preview |
| LMArena Text: expert | 1357 | 1422 | Hy3 preview |
| LMArena Text: hard prompts in English | 1369 | 1416 | Hy3 preview |
| Neither model has a result on the other 16 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | GLM 4.7 Flash | Hy3 preview | Leader |
|---|---|---|---|
| OTIS Mock AIME | 58.3% | Not measured | Only GLM 4.7 Flash measured |
| LMArena Text: math | 1355 | 1424 | Hy3 preview |
| LMArena Text: mathematical work | 1369 | 1442 | Hy3 preview |
| Neither model has a result on the other 10 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
| Benchmark | GLM 4.7 Flash | Hy3 preview | Leader |
|---|---|---|---|
| LMArena Text: science | 1358 | 1416 | Hy3 preview |
| LMArena Text: medicine and healthcare | 1343 | 1405 | Hy3 preview |
| LMArena Text: legal and government | 1357 | 1430 | Hy3 preview |
| LMArena Text: business and finance | 1343 | 1397 | Hy3 preview |
| Neither model has a result on the other 14 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
| Benchmark | GLM 4.7 Flash | Hy3 preview | Leader |
|---|---|---|---|
| Design Arena: websites | 1194 | Not measured | Only GLM 4.7 Flash measured |
| Design Arena: UI components | 1208 | Not measured | Only GLM 4.7 Flash measured |
| Design Arena: game dev | 1135 | Not measured | Only GLM 4.7 Flash measured |
| Design Arena: data viz | 1123 | Not measured | Only GLM 4.7 Flash measured |
| Design Arena: 3D | 1121 | Not measured | Only GLM 4.7 Flash measured |
| Design Arena: all code categories | 1178 | Not measured | Only GLM 4.7 Flash measured |
| Design Arena: SVG | 1038 | Not measured | Only GLM 4.7 Flash measured |
| Neither model has a result on the other 2 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index015
| Benchmark | GLM 4.7 Flash | Hy3 preview | Leader |
|---|---|---|---|
| IFBench | 60.8% | 63.1% | Hy3 preview |
| Long context reasoning (AA-LCR) | 41.7% | 64.7% | Hy3 preview |
| LMArena Text | 1351 | 1401 | Hy3 preview |
| LMArena Text: instruction following | 1327 | 1383 | Hy3 preview |
| LMArena Text: creative writing | 1296 | 1339 | Hy3 preview |
| LMArena Text: English | 1373 | 1409 | Hy3 preview |
| LMArena Text: non-English | 1330 | 1390 | Hy3 preview |
| LMArena Text: Chinese | 1403 | 1435 | Hy3 preview |
| LMArena Text: German | 1336 | Not measured | Only GLM 4.7 Flash measured |
| LMArena Text: Korean | 1283 | Not measured | Only GLM 4.7 Flash measured |
| LMArena Text: French | 1332 | 1425 | Hy3 preview |
| LMArena Text: Spanish | 1350 | 1399 | Hy3 preview |
| LMArena Text: Russian | 1333 | 1399 | Hy3 preview |
| LMArena Text: Polish | 1291 | Not measured | Only GLM 4.7 Flash measured |
| LMArena Text: multi-turn | 1342 | 1411 | Hy3 preview |
| LMArena Text: longer queries | 1345 | 1402 | Hy3 preview |
| LMArena Text: writing and language | 1320 | 1370 | Hy3 preview |
| LMArena Text: entertainment and media | 1304 | 1352 | Hy3 preview |
| Neither model has a result on the other 28 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GLM 4.7 Flash | Hy3 preview | 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
GLM 4.7 Flash costs less
Repository review
50K input + 3K output tokens
GLM 4.7 Flash costs less
Agent loop
200K input + 10K output tokens
GLM 4.7 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.
| Specification | Z.ai GLM 4.7 Flash | Tencent Hy3 preview |
|---|---|---|
| Maker | Z.ai | Tencent |
| Weights | Open weights | Open weights |
| Released | January 19, 2026 | April 22, 2026 |
| Context window | 200K | 262K |
| Input $/Mtok | $0.0653 | $0.18 |
| Output $/Mtok | $0.40 | $0.60 |
| 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 GLM 4.7 Flash better than Hy3 preview for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick GLM 4.7 Flash for output tokens 1.5x cheaper ($0.40 vs $0.60 per million) and work measured by Terminal-Bench Hard (Artificial Analysis run) or LMArena Text: coding, where it leads.
Which is cheaper, GLM 4.7 Flash or Hy3 preview?
GLM 4.7 Flash costs $0.0653 per million input tokens and $0.40 per million output tokens; Hy3 preview costs $0.18 and $0.60. Cost per solved task is known for neither of them.
Which has the bigger context window, GLM 4.7 Flash or Hy3 preview?
Hy3 preview, with 262K tokens against 200K.
How many benchmarks is the GLM 4.7 Flash vs Hy3 preview comparison based on?
31 benchmarks have a result for both models as of October 21, 2026. GLM 4.7 Flash's coding index rests on 0 independent benchmarks and Hy3 preview's on 1. Where only one model has a result, the page says so instead of naming a winner.