Gemma 3 12B vs GPT-4o-mini (2024-07-18)
Gemma 3 12B costs 3.7x less per output token than GPT-4o-mini (2024-07-18) ($0.1625 vs $0.60 per million). Updated October 22, 2026; 24 benchmarks have a result for both.
Gemma 3 12B or GPT-4o-mini (2024-07-18): 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 3.7x cheaper ($0.1625 vs $0.60 per million) and a larger context window (131K against 128K tokens).
- Shared results
- 24
- Gemma 3 12B only
- 19
- GPT-4o-mini (2024-07-18) only
- 22
- 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
Intelligence
AA Intelligence Index
No call
Only Gemma 3 12B has a AA Intelligence Index result, 3.8.
Directional only
Agentic work
AA Agentic Index
No call
Only Gemma 3 12B has a AA Agentic Index result, 0.1.
Directional only
Long documents
Listed context window
Gemma 3 12B
Gemma 3 12B takes 131K tokens against 128K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Gemma 3 12B
Gemma 3 12B $0.0002, GPT-4o-mini (2024-07-18) $0.0004 per turn.
List prices
At a glance
- Coding index
- Gemma 3 12B not benchmarked, GPT-4o-mini (2024-07-18) not benchmarked.
- Widest category
- Knowledge, 13.3 points apart on average over 4 shared benchmarks.
- Widest single gap
- LMArena Text: business and finance: Gemma 3 12B 1376 against GPT-4o-mini (2024-07-18) 1265, 18.6 points apart on a 0 to 100 scale.
- Output price
- Gemma 3 12B $0.1625 against GPT-4o-mini (2024-07-18) $0.60 per million output tokens, 3.7x apart.
- Context window
- Gemma 3 12B takes 131K tokens, GPT-4o-mini (2024-07-18) 128K.
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 | GPT-4o-mini (2024-07-18) | Ties |
|---|---|---|---|---|
| Coding | 1 | 1 | ||
| Reasoning | 4 | 0 | ||
| Math | 3 | 0 | ||
| Knowledge | 4 | 0 | ||
| General | 11 | 0 | ||
| All shared boards | 23 | 1 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index11
| Benchmark | Gemma 3 12B | GPT-4o-mini (2024-07-18) | Leader |
|---|---|---|---|
| Aider polyglot | Not measured | 3.6% | Only GPT-4o-mini (2024-07-18) 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% | Not measured | Only Gemma 3 12B measured |
| WeirdML | Not measured | 11.8% | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: coding | 1281 | 1290 | GPT-4o-mini (2024-07-18) |
| LMArena Text: software and IT | 1313 | 1291 | Gemma 3 12B |
| SciCode | 0.0% | Not measured | Only Gemma 3 12B measured |
| Neither model has a result on the other 34 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Gemma 3 12B | GPT-4o-mini (2024-07-18) | Leader |
|---|---|---|---|
| AA Intelligence Index | 3.8 | Not measured | Only Gemma 3 12B measured |
| Epoch Capabilities Index | 123.5 | Not measured | Only Gemma 3 12B measured |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index4 boards, none shared
| Benchmark | Gemma 3 12B | GPT-4o-mini (2024-07-18) | Leader |
|---|---|---|---|
| AA Agentic Index | 0.1 | Not measured | Only Gemma 3 12B measured |
| τ²-Bench Telecom | 10.8% | Not measured | Only Gemma 3 12B measured |
| BALROG | Not measured | 17.4% | Only GPT-4o-mini (2024-07-18) measured |
| Berkeley Function Calling Leaderboard (BFCL) | 30.4% | Not measured | Only Gemma 3 12B measured |
| Neither model has a result on the other 38 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | Gemma 3 12B | GPT-4o-mini (2024-07-18) | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 34.9% | Not measured | Only Gemma 3 12B measured |
| Humanity's Last Exam (Artificial Analysis run) | 4.2% | Not measured | Only Gemma 3 12B measured |
| GPQA Diamond (Epoch AI run) | 39.5% | 37.7% | Gemma 3 12B |
| SimpleBench | Not measured | 10.7% | Only GPT-4o-mini (2024-07-18) measured |
| Chess puzzles | Not measured | 3.0% | Only GPT-4o-mini (2024-07-18) measured |
| Mystery game puzzles | Not measured | 12.0% | Only GPT-4o-mini (2024-07-18) measured |
| VPCT | Not measured | 34.0% | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: hard prompts | 1309 | 1267 | Gemma 3 12B |
| LMArena Text: expert | 1248 | 1234 | Gemma 3 12B |
| LMArena Text: hard prompts in English | 1318 | 1278 | Gemma 3 12B |
| Neither model has a result on the other 13 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | Gemma 3 12B | GPT-4o-mini (2024-07-18) | 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% | 6.9% | Gemma 3 12B |
| FrontierMath, tiers 1 to 3 | Not measured | 0.7% | Only GPT-4o-mini (2024-07-18) measured |
| MATH level 5 | Not measured | 52.6% | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: math | 1307 | 1267 | Gemma 3 12B |
| LMArena Text: mathematical work | 1336 | 1271 | Gemma 3 12B |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index40
| Benchmark | Gemma 3 12B | GPT-4o-mini (2024-07-18) | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | 59.5% | Not measured | Only Gemma 3 12B measured |
| SimpleQA Verified | Not measured | 8.3% | Only GPT-4o-mini (2024-07-18) measured |
| MMLU | Not measured | 76.7% | Only GPT-4o-mini (2024-07-18) measured |
| GeoBench | Not measured | 64.0% | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: science | 1366 | 1286 | Gemma 3 12B |
| LMArena Text: medicine and healthcare | 1334 | 1282 | Gemma 3 12B |
| LMArena Text: legal and government | 1381 | 1305 | Gemma 3 12B |
| LMArena Text: business and finance | 1376 | 1265 | Gemma 3 12B |
| Neither model has a result on the other 12 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemma 3 12B | GPT-4o-mini (2024-07-18) | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index110
| Benchmark | Gemma 3 12B | GPT-4o-mini (2024-07-18) | Leader |
|---|---|---|---|
| IFBench | 36.7% | Not measured | Only Gemma 3 12B measured |
| Long context reasoning (AA-LCR) | 8.3% | Not measured | Only Gemma 3 12B measured |
| DTBench | 48.8% | Not measured | Only Gemma 3 12B measured |
| LMCA | 4.5% | Not measured | Only Gemma 3 12B measured |
| Creative writing (Lech Mazur) | Not measured | 67.2% | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text | 1334 | 1286 | Gemma 3 12B |
| LMArena Text: instruction following | 1299 | 1258 | Gemma 3 12B |
| LMArena Text: creative writing | 1331 | 1268 | Gemma 3 12B |
| LMArena Vision | Not measured | 1066 | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: English | 1349 | 1301 | Gemma 3 12B |
| LMArena Text: non-English | 1318 | 1266 | Gemma 3 12B |
| LMArena Text: Chinese | Not measured | 1265 | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: German | 1370 | 1272 | Gemma 3 12B |
| LMArena Text: Japanese | Not measured | 1216 | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: Korean | Not measured | 1195 | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: French | Not measured | 1297 | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: Spanish | Not measured | 1276 | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: Russian | 1335 | 1275 | Gemma 3 12B |
| LMArena Text: Polish | Not measured | 1244 | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Text: multi-turn | 1334 | 1285 | Gemma 3 12B |
| LMArena Text: longer queries | 1317 | 1289 | Gemma 3 12B |
| LMArena Text: writing and language | 1307 | 1282 | Gemma 3 12B |
| LMArena Text: entertainment and media | 1297 | 1253 | Gemma 3 12B |
| LMArena Vision: English | Not measured | 1072 | Only GPT-4o-mini (2024-07-18) measured |
| LMArena Vision: Chinese | Not measured | 1033 | Only GPT-4o-mini (2024-07-18) measured |
| Neither model has a result on the other 23 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemma 3 12B | GPT-4o-mini (2024-07-18) | 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 | OpenAI GPT-4o-mini (2024-07-18) |
|---|---|---|
| Maker | OpenAI | |
| Weights | Open weights | No public weights listed |
| Released | March 13, 2025 | July 18, 2024 |
| Context window | 131K | 128K |
| Input $/Mtok | $0.075 | $0.15 |
| Output $/Mtok | $0.1625 | $0.60 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Gemma 3 12B better than GPT-4o-mini (2024-07-18) for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Gemma 3 12B for output tokens 3.7x cheaper ($0.1625 vs $0.60 per million) and a larger context window (131K against 128K tokens).
Which is cheaper, Gemma 3 12B or GPT-4o-mini (2024-07-18)?
Gemma 3 12B costs $0.075 per million input tokens and $0.1625 per million output tokens; GPT-4o-mini (2024-07-18) costs $0.15 and $0.60. Cost per solved task is known for neither of them.
Which has the bigger context window, Gemma 3 12B or GPT-4o-mini (2024-07-18)?
Gemma 3 12B, with 131K tokens against 128K.
How many benchmarks is the Gemma 3 12B vs GPT-4o-mini (2024-07-18) comparison based on?
24 benchmarks have a result for both models as of October 22, 2026. Gemma 3 12B's coding index rests on 0 independent benchmarks and GPT-4o-mini (2024-07-18)'s on 0. Where only one model has a result, the page says so instead of naming a winner.