Gemma 3 12B vs GLM 5V Turbo
GLM 5V Turbo leads Gemma 3 12B on the Artificial Analysis Intelligence Index, 23.5 to 3.8, and Gemma 3 12B costs 24.6x less per output token ($0.1625 vs $4.00 per million). Updated October 22, 2026; 29 benchmarks have a result for both.
Gemma 3 12B or GLM 5V Turbo: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Gemma 3 12B for output tokens 24.6x cheaper ($0.1625 vs $4.00 per million) and work measured by Terminal-Bench Hard (Artificial Analysis run) or LMArena Text: coding, where it leads.
- Shared results
- 29
- Gemma 3 12B only
- 14
- GLM 5V Turbo only
- 36
- 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
GLM 5V Turbo
GLM 5V Turbo leads, 23.5 against 3.8.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Gemma 3 12B has a AA Agentic Index result, 0.1.
Directional only
Web design
Design Arena: websites
No call
Only GLM 5V Turbo has a Design Arena: websites result, 1230.
Directional only
Long documents
Listed context window
GLM 5V Turbo
GLM 5V Turbo takes 203K tokens against 131K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Gemma 3 12B
Gemma 3 12B $0.0002, GLM 5V Turbo $0.0032 per turn.
List prices
At a glance
- Coding index
- Gemma 3 12B not benchmarked, GLM 5V Turbo not benchmarked.
- Widest category
- Agentic work, 87.7 points apart on average over 1 shared benchmark.
- Widest single gap
- τ²-Bench Telecom: Gemma 3 12B 10.8% against GLM 5V Turbo 98.5%, 87.7 points apart on a 0 to 100 scale.
- Output price
- Gemma 3 12B $0.1625 against GLM 5V Turbo $4.00 per million output tokens, 24.6x apart.
- Context window
- Gemma 3 12B takes 131K tokens, GLM 5V Turbo 203K.
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 | Gemma 3 12B | Bars | GLM 5V Turbo | Ties |
|---|---|---|---|---|
| Coding | 0 | 3 | ||
| Intelligence | 0 | 1 | ||
| Agentic work | 0 | 1 | ||
| Reasoning | 0 | 5 | ||
| Math | 0 | 2 | ||
| Knowledge | 0 | 4 | ||
| General | 0 | 13 | ||
| All shared boards | 0 | 29 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index03
| Benchmark | Gemma 3 12B | GLM 5V Turbo | Leader |
|---|---|---|---|
| LMArena WebDev | Not measured | 1402 | Only GLM 5V Turbo measured |
| AA Coding Index | 5.8 | Not measured | Only Gemma 3 12B measured |
| LiveCodeBench (Artificial Analysis run) | 13.7% | Not measured | Only Gemma 3 12B measured |
| SciCode | 16.4% | Not measured | Only Gemma 3 12B measured |
| Terminal-Bench Hard (Artificial Analysis run) | 0.8% | 32.6% | GLM 5V Turbo |
| LMArena Text: coding | 1281 | 1466 | GLM 5V Turbo |
| LMArena Text: software and IT | 1313 | 1460 | GLM 5V Turbo |
| LMArena WebDev: React | Not measured | 1386 | Only GLM 5V Turbo measured |
| LMArena WebDev: HTML | Not measured | 1356 | Only GLM 5V Turbo measured |
| LMArena WebDev: from an image | Not measured | 1425 | Only GLM 5V Turbo measured |
| SciCode | 0.0% | Not measured | Only Gemma 3 12B measured |
| Neither model has a result on the other 28 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Gemma 3 12B | GLM 5V Turbo | Leader |
|---|---|---|---|
| AA Intelligence Index | 3.8 | 23.5 | GLM 5V Turbo |
| Epoch Capabilities Index | 123.5 | Not measured | Only Gemma 3 12B measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Gemma 3 12B | GLM 5V Turbo | Leader |
|---|---|---|---|
| AA Agentic Index | 0.1 | Not measured | Only Gemma 3 12B measured |
| Design Arena: agents building web apps | Not measured | 1124 | Only GLM 5V Turbo measured |
| Design Arena: agents building full-stack apps | Not measured | 1149 | Only GLM 5V Turbo measured |
| Design Arena: agents building mobile apps | Not measured | 1134 | Only GLM 5V Turbo measured |
| Design Arena: agents building Android apps | Not measured | 1267 | Only GLM 5V Turbo measured |
| Design Arena: agents building Godot games | Not measured | 957 | Only GLM 5V Turbo measured |
| Design Arena: agents building games | Not measured | 1102 | Only GLM 5V Turbo measured |
| Design Arena: HTML slides | Not measured | 1070 | Only GLM 5V Turbo measured |
| Design Arena: PowerPoint slides | Not measured | 1127 | Only GLM 5V Turbo measured |
| τ²-Bench Telecom | 10.8% | 98.5% | GLM 5V Turbo |
| Berkeley Function Calling Leaderboard (BFCL) | 30.4% | Not measured | Only Gemma 3 12B measured |
| Neither model has a result on the other 30 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index05
| Benchmark | Gemma 3 12B | GLM 5V Turbo | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 34.9% | 80.9% | GLM 5V Turbo |
| Humanity's Last Exam (Artificial Analysis run) | 4.2% | 17.1% | GLM 5V Turbo |
| GPQA Diamond (Epoch AI run) | 39.5% | Not measured | Only Gemma 3 12B measured |
| LMArena Text: hard prompts | 1309 | 1443 | GLM 5V Turbo |
| LMArena Text: expert | 1248 | 1453 | GLM 5V Turbo |
| LMArena Text: hard prompts in English | 1318 | 1456 | GLM 5V Turbo |
| LMArena Vision: diagrams | Not measured | 1286 | Only GLM 5V Turbo measured |
| LMArena Vision: homework | Not measured | 1273 | Only GLM 5V Turbo measured |
| Neither model has a result on the other 14 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Gemma 3 12B | GLM 5V Turbo | Leader |
|---|---|---|---|
| AA Math Index | 18.3 | Not measured | Only Gemma 3 12B measured |
| AIME 2025 (Artificial Analysis run) | 18.3% | Not measured | Only Gemma 3 12B measured |
| OTIS Mock AIME | 16.7% | Not measured | Only Gemma 3 12B measured |
| LMArena Text: math | 1307 | 1442 | GLM 5V Turbo |
| LMArena Text: mathematical work | 1336 | 1447 | GLM 5V Turbo |
| Neither model has a result on the other 8 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index04
| Benchmark | Gemma 3 12B | GLM 5V Turbo | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | 59.5% | Not measured | Only Gemma 3 12B measured |
| LMArena Text: science | 1366 | 1448 | GLM 5V Turbo |
| LMArena Text: medicine and healthcare | 1334 | 1433 | GLM 5V Turbo |
| LMArena Text: legal and government | 1381 | 1436 | GLM 5V Turbo |
| LMArena Text: business and finance | 1376 | 1435 | GLM 5V Turbo |
| Neither model has a result on the other 13 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index8 boards, none shared
| Benchmark | Gemma 3 12B | GLM 5V Turbo | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1230 | Only GLM 5V Turbo measured |
| Design Arena: UI components | Not measured | 1210 | Only GLM 5V Turbo measured |
| Design Arena: game dev | Not measured | 1218 | Only GLM 5V Turbo measured |
| Design Arena: data viz | Not measured | 1202 | Only GLM 5V Turbo measured |
| Design Arena: 3D | Not measured | 1208 | Only GLM 5V Turbo measured |
| Design Arena: all code categories | Not measured | 1228 | Only GLM 5V Turbo measured |
| Design Arena: SVG | Not measured | 1151 | Only GLM 5V Turbo measured |
| Design Arena: ASCII art | Not measured | 1117 | Only GLM 5V Turbo measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index013
| Benchmark | Gemma 3 12B | GLM 5V Turbo | Leader |
|---|---|---|---|
| IFBench | 36.7% | 61.1% | GLM 5V Turbo |
| Long context reasoning (AA-LCR) | 8.3% | 70.3% | GLM 5V Turbo |
| DTBench | 48.8% | Not measured | Only Gemma 3 12B measured |
| LMCA | 4.5% | Not measured | Only Gemma 3 12B measured |
| LMArena Text | 1334 | 1437 | GLM 5V Turbo |
| LMArena Text: instruction following | 1299 | 1423 | GLM 5V Turbo |
| LMArena Text: creative writing | 1331 | 1417 | GLM 5V Turbo |
| LMArena Vision | Not measured | 1264 | Only GLM 5V Turbo measured |
| LMArena Text: English | 1349 | 1454 | GLM 5V Turbo |
| LMArena Text: non-English | 1318 | 1420 | GLM 5V Turbo |
| LMArena Text: Chinese | Not measured | 1491 | Only GLM 5V Turbo measured |
| LMArena Text: German | 1370 | 1422 | GLM 5V Turbo |
| LMArena Text: Korean | Not measured | 1397 | Only GLM 5V Turbo measured |
| LMArena Text: French | Not measured | 1443 | Only GLM 5V Turbo measured |
| LMArena Text: Spanish | Not measured | 1452 | Only GLM 5V Turbo measured |
| LMArena Text: Russian | 1335 | 1432 | GLM 5V Turbo |
| LMArena Text: Polish | Not measured | 1432 | Only GLM 5V Turbo measured |
| LMArena Text: multi-turn | 1334 | 1431 | GLM 5V Turbo |
| LMArena Text: longer queries | 1317 | 1438 | GLM 5V Turbo |
| LMArena Text: writing and language | 1307 | 1420 | GLM 5V Turbo |
| LMArena Text: entertainment and media | 1297 | 1407 | GLM 5V Turbo |
| LMArena Vision: English | Not measured | 1274 | Only GLM 5V Turbo measured |
| LMArena Vision: Chinese | Not measured | 1329 | Only GLM 5V Turbo measured |
| LMArena Vision: reading text | Not measured | 1270 | Only GLM 5V Turbo measured |
| LMArena Vision: humor | Not measured | 1268 | Only GLM 5V Turbo measured |
| LMArena Vision: captioning | Not measured | 1264 | Only GLM 5V Turbo measured |
| LMArena Vision: recognizing things | Not measured | 1241 | Only GLM 5V Turbo measured |
| LMArena Vision: creative writing | Not measured | 1252 | Only GLM 5V Turbo measured |
| LMArena Document | Not measured | 1416 | Only GLM 5V Turbo measured |
| Neither model has a result on the other 17 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemma 3 12B | GLM 5V Turbo | 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
Gemma 3 12B costs less
Repository review
50K input + 3K output tokens
Gemma 3 12B costs less
Agent loop
200K input + 10K output tokens
Gemma 3 12B 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 | Google Gemma 3 12B | Z.ai GLM 5V Turbo |
|---|---|---|
| Maker | Z.ai | |
| Weights | Open weights | No public weights listed |
| Released | March 13, 2025 | April 1, 2026 |
| Context window | 131K | 203K |
| Input $/Mtok | $0.075 | $1.20 |
| Output $/Mtok | $0.1625 | $4.00 |
| 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 Gemma 3 12B better than GLM 5V Turbo for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Gemma 3 12B for output tokens 24.6x cheaper ($0.1625 vs $4.00 per million) and work measured by Terminal-Bench Hard (Artificial Analysis run) or LMArena Text: coding, where it leads.
Which is cheaper, Gemma 3 12B or GLM 5V Turbo?
Gemma 3 12B costs $0.075 per million input tokens and $0.1625 per million output tokens; GLM 5V Turbo costs $1.20 and $4.00. Cost per solved task is known for neither of them.
Which has the bigger context window, Gemma 3 12B or GLM 5V Turbo?
GLM 5V Turbo, with 203K tokens against 131K.
How many benchmarks is the Gemma 3 12B vs GLM 5V Turbo comparison based on?
29 benchmarks have a result for both models as of October 22, 2026. Gemma 3 12B's coding index rests on 0 independent benchmarks and GLM 5V Turbo's on 1. Where only one model has a result, the page says so instead of naming a winner.