Gemma 3 4B vs GPT-4o (2024-05-13)
GPT-4o (2024-05-13) leads Gemma 3 4B on the Artificial Analysis Intelligence Index, 7.3 to 4.8, and Gemma 3 4B costs 200.0x less per output token ($0.075 vs $15.00 per million). Updated October 22, 2026; 30 benchmarks have a result for both.
Gemma 3 4B or GPT-4o (2024-05-13): which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Gemma 3 4B for output tokens 200.0x cheaper ($0.075 vs $15.00 per million) and a larger context window (131K against 128K tokens).
- Shared results
- 30
- Gemma 3 4B only
- 12
- GPT-4o (2024-05-13) only
- 14
- Like-for-like categories
- 6 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
GPT-4o (2024-05-13)
GPT-4o (2024-05-13) leads, 7.3 against 4.8.
Same board, both measured
Reasoning
GPQA Diamond
No call
Only GPT-4o (2024-05-13) has a GPQA Diamond result, 50.2%.
Directional only
Long documents
Listed context window
Gemma 3 4B
Gemma 3 4B takes 131K tokens against 128K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Gemma 3 4B
Gemma 3 4B $0.0001, GPT-4o (2024-05-13) $0.013 per turn.
List prices
At a glance
- Coding index
- Gemma 3 4B not benchmarked, GPT-4o (2024-05-13) not benchmarked.
- Widest category
- Coding, 14.9 points apart on average over 4 shared benchmarks.
- Widest single gap
- MMLU Pro (Artificial Analysis run): Gemma 3 4B 41.7% against GPT-4o (2024-05-13) 74.0%, 32.3 points apart on a 0 to 100 scale.
- Output price
- Gemma 3 4B $0.075 against GPT-4o (2024-05-13) $15.00 per million output tokens, 200.0x apart.
- Context window
- Gemma 3 4B takes 131K tokens, GPT-4o (2024-05-13) 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 4B | Bars | GPT-4o (2024-05-13) | Ties |
|---|---|---|---|---|
| Coding | 0 | 4 | ||
| Intelligence | 0 | 1 | ||
| Reasoning | 1 | 5 | ||
| Math | 1 | 2 | ||
| Knowledge | 4 | 1 | ||
| General | 1 | 10 | ||
| All shared boards | 7 | 23 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index04
| Benchmark | Gemma 3 4B | GPT-4o (2024-05-13) | Leader |
|---|---|---|---|
| AA Coding Index | 2.7 | 24.2 | GPT-4o (2024-05-13) |
| LiveCodeBench (Artificial Analysis run) | 11.2% | 33.4% | GPT-4o (2024-05-13) |
| Terminal-Bench Hard (Artificial Analysis run) | 0.8% | Not measured | Only Gemma 3 4B measured |
| LMArena Text: coding | 1230 | 1297 | GPT-4o (2024-05-13) |
| LMArena Text: software and IT | 1269 | 1297 | GPT-4o (2024-05-13) |
| SciCode | 0.0% | Not measured | Only Gemma 3 4B measured |
| Neither model has a result on the other 37 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Gemma 3 4B | GPT-4o (2024-05-13) | Leader |
|---|---|---|---|
| AA Intelligence Index | 4.8 | 7.3 | GPT-4o (2024-05-13) |
| Epoch Capabilities Index | 115.9 | Not measured | Only Gemma 3 4B measured |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index4 boards, none shared
| Benchmark | Gemma 3 4B | GPT-4o (2024-05-13) | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 5.0% | Not measured | Only Gemma 3 4B measured |
| METR time horizons | Not measured | 40.8% | Only GPT-4o (2024-05-13) measured |
| BALROG | Not measured | 32.3% | Only GPT-4o (2024-05-13) measured |
| Berkeley Function Calling Leaderboard (BFCL) | 19.6% | Not measured | Only Gemma 3 4B measured |
| Neither model has a result on the other 38 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index15
| Benchmark | Gemma 3 4B | GPT-4o (2024-05-13) | Leader |
|---|---|---|---|
| GPQA Diamond | Not measured | 50.2% | Only GPT-4o (2024-05-13) measured |
| GPQA Diamond (Artificial Analysis run) | 29.1% | 52.6% | GPT-4o (2024-05-13) |
| Humanity's Last Exam (Artificial Analysis run) | 5.3% | 1.8% | Gemma 3 4B |
| GPQA Diamond (Epoch AI run) | 23.2% | 48.9% | GPT-4o (2024-05-13) |
| LMArena Text: hard prompts | 1253 | 1281 | GPT-4o (2024-05-13) |
| LMArena Text: expert | 1223 | 1250 | GPT-4o (2024-05-13) |
| LMArena Text: hard prompts in English | 1253 | 1293 | GPT-4o (2024-05-13) |
| Kagi LLM Benchmark | 25.2% | Not measured | Only Gemma 3 4B measured |
| Neither model has a result on the other 15 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index12
| Benchmark | Gemma 3 4B | GPT-4o (2024-05-13) | Leader |
|---|---|---|---|
| AA Math Index | 12.7 | Not measured | Only Gemma 3 4B measured |
| AIME 2025 (Artificial Analysis run) | 12.7% | Not measured | Only Gemma 3 4B measured |
| OTIS Mock AIME | 7.5% | 6.3% | Gemma 3 4B |
| MATH level 5 | Not measured | 51.0% | Only GPT-4o (2024-05-13) measured |
| LMArena Text: math | 1239 | 1284 | GPT-4o (2024-05-13) |
| LMArena Text: mathematical work | 1256 | 1284 | GPT-4o (2024-05-13) |
| Neither model has a result on the other 8 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index41
| Benchmark | Gemma 3 4B | GPT-4o (2024-05-13) | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | 41.7% | 74.0% | GPT-4o (2024-05-13) |
| MMLU | Not measured | 84.2% | Only GPT-4o (2024-05-13) measured |
| LMArena Text: science | 1306 | 1295 | Gemma 3 4B |
| LMArena Text: medicine and healthcare | 1342 | 1278 | Gemma 3 4B |
| LMArena Text: legal and government | 1327 | 1314 | Gemma 3 4B |
| LMArena Text: business and finance | 1311 | 1262 | Gemma 3 4B |
| Neither model has a result on the other 14 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemma 3 4B | GPT-4o (2024-05-13) | 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 4B | GPT-4o (2024-05-13) | Leader |
|---|---|---|---|
| IFBench | 28.3% | Not measured | Only Gemma 3 4B measured |
| Long context reasoning (AA-LCR) | 6.7% | Not measured | Only Gemma 3 4B measured |
| DTBench | 50.9% | Not measured | Only Gemma 3 4B measured |
| LMCA | 2.8% | Not measured | Only Gemma 3 4B measured |
| LMArena Text | 1291 | 1300 | GPT-4o (2024-05-13) |
| LMArena Text: instruction following | 1239 | 1278 | GPT-4o (2024-05-13) |
| LMArena Text: creative writing | 1270 | 1291 | GPT-4o (2024-05-13) |
| LMArena Vision | Not measured | 1137 | Only GPT-4o (2024-05-13) measured |
| LMArena Text: English | 1300 | 1312 | GPT-4o (2024-05-13) |
| LMArena Text: non-English | 1273 | 1283 | GPT-4o (2024-05-13) |
| LMArena Text: Chinese | Not measured | 1277 | Only GPT-4o (2024-05-13) measured |
| LMArena Text: German | 1281 | 1283 | GPT-4o (2024-05-13) |
| LMArena Text: Japanese | Not measured | 1257 | Only GPT-4o (2024-05-13) measured |
| LMArena Text: Korean | Not measured | 1234 | Only GPT-4o (2024-05-13) measured |
| LMArena Text: French | Not measured | 1304 | Only GPT-4o (2024-05-13) measured |
| LMArena Text: Spanish | Not measured | 1292 | Only GPT-4o (2024-05-13) measured |
| LMArena Text: Russian | 1294 | 1285 | Gemma 3 4B |
| LMArena Text: Polish | Not measured | 1365 | Only GPT-4o (2024-05-13) measured |
| LMArena Text: multi-turn | 1255 | 1302 | GPT-4o (2024-05-13) |
| LMArena Text: longer queries | 1273 | 1288 | GPT-4o (2024-05-13) |
| LMArena Text: writing and language | 1288 | 1298 | GPT-4o (2024-05-13) |
| LMArena Text: entertainment and media | 1233 | 1289 | GPT-4o (2024-05-13) |
| LMArena Vision: English | Not measured | 1147 | Only GPT-4o (2024-05-13) measured |
| LMArena Vision: Chinese | Not measured | 1132 | Only GPT-4o (2024-05-13) 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
| Benchmark | Gemma 3 4B | GPT-4o (2024-05-13) | 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 4B costs less
Repository review
50K input + 3K output tokens
Gemma 3 4B costs less
Agent loop
200K input + 10K output tokens
Gemma 3 4B 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 4B | OpenAI GPT-4o (2024-05-13) |
|---|---|---|
| Maker | OpenAI | |
| Weights | Open weights | No public weights listed |
| Released | March 13, 2025 | May 13, 2024 |
| Context window | 131K | 128K |
| Input $/Mtok | $0.045 | $5.00 |
| Output $/Mtok | $0.075 | $15.00 |
| 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 4B better than GPT-4o (2024-05-13) for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Gemma 3 4B for output tokens 200.0x cheaper ($0.075 vs $15.00 per million) and a larger context window (131K against 128K tokens).
Which is cheaper, Gemma 3 4B or GPT-4o (2024-05-13)?
Gemma 3 4B costs $0.045 per million input tokens and $0.075 per million output tokens; GPT-4o (2024-05-13) costs $5.00 and $15.00. Cost per solved task is known for neither of them.
Which has the bigger context window, Gemma 3 4B or GPT-4o (2024-05-13)?
Gemma 3 4B, with 131K tokens against 128K.
How many benchmarks is the Gemma 3 4B vs GPT-4o (2024-05-13) comparison based on?
30 benchmarks have a result for both models as of October 22, 2026. Gemma 3 4B's coding index rests on 0 independent benchmarks and GPT-4o (2024-05-13)'s on 0. Where only one model has a result, the page says so instead of naming a winner.