Gemma 3 4B vs GPT-4 Turbo
GPT-4 Turbo leads Gemma 3 4B on the Artificial Analysis Intelligence Index, 7.0 to 4.8, and Gemma 3 4B costs 400.0x less per output token ($0.075 vs $30.00 per million). Updated October 22, 2026; 30 benchmarks have a result for both.
Gemma 3 4B or GPT-4 Turbo: 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 400.0x cheaper ($0.075 vs $30.00 per million) and a larger context window (131K against 128K tokens).
- Shared results
- 30
- Gemma 3 4B only
- 12
- GPT-4 Turbo only
- 12
- 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-4 Turbo
GPT-4 Turbo leads, 7.0 against 4.8.
Same board, both measured
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-4 Turbo $0.025 per turn.
List prices
At a glance
- Coding index
- Gemma 3 4B not benchmarked, GPT-4 Turbo not benchmarked.
- Widest category
- Knowledge, 15.6 points apart on average over 5 shared benchmarks.
- Widest single gap
- MMLU Pro (Artificial Analysis run): Gemma 3 4B 41.7% against GPT-4 Turbo 69.4%, 27.7 points apart on a 0 to 100 scale.
- Output price
- Gemma 3 4B $0.075 against GPT-4 Turbo $30.00 per million output tokens, 400.0x apart.
- Context window
- Gemma 3 4B takes 131K tokens, GPT-4 Turbo 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-4 Turbo | Ties |
|---|---|---|---|---|
| Coding | 1 | 3 | ||
| Intelligence | 0 | 2 | ||
| Reasoning | 2 | 2 | ||
| Math | 0 | 2 | ||
| Knowledge | 4 | 1 | ||
| General | 8 | 5 | ||
| All shared boards | 15 | 15 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index13
| Benchmark | Gemma 3 4B | GPT-4 Turbo | Leader |
|---|---|---|---|
| AA Coding Index | 2.7 | 21.5 | GPT-4 Turbo |
| LiveCodeBench (Artificial Analysis run) | 11.2% | 29.1% | GPT-4 Turbo |
| Terminal-Bench Hard (Artificial Analysis run) | 0.8% | Not measured | Only Gemma 3 4B measured |
| LMArena Text: coding | 1230 | 1268 | GPT-4 Turbo |
| LMArena Text: software and IT | 1269 | 1267 | Gemma 3 4B |
| SciCode | 0.0% | Not measured | Only Gemma 3 4B measured |
| Neither model has a result on the other 33 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Gemma 3 4B | GPT-4 Turbo | Leader |
|---|---|---|---|
| AA Intelligence Index | 4.8 | 7.0 | GPT-4 Turbo |
| Epoch Capabilities Index | 115.9 | 127.3 | GPT-4 Turbo |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Gemma 3 4B | GPT-4 Turbo | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 5.0% | Not measured | Only Gemma 3 4B measured |
| Berkeley Function Calling Leaderboard (BFCL) | 19.6% | Not measured | Only Gemma 3 4B measured |
| Neither model has a result on the other 39 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index22
| Benchmark | Gemma 3 4B | GPT-4 Turbo | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 29.1% | Not measured | Only Gemma 3 4B measured |
| Humanity's Last Exam (Artificial Analysis run) | 5.3% | 3.1% | Gemma 3 4B |
| GPQA Diamond (Epoch AI run) | 23.2% | Not measured | Only Gemma 3 4B measured |
| LMArena Text: hard prompts | 1253 | 1251 | Gemma 3 4B |
| LMArena Text: expert | 1223 | 1223 | GPT-4 Turbo |
| LMArena Text: hard prompts in English | 1253 | 1266 | GPT-4 Turbo |
| 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 index02
| Benchmark | Gemma 3 4B | GPT-4 Turbo | 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% | Not measured | Only Gemma 3 4B measured |
| LMArena Text: math | 1239 | 1272 | GPT-4 Turbo |
| LMArena Text: mathematical work | 1256 | 1268 | GPT-4 Turbo |
| 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-4 Turbo | Leader |
|---|---|---|---|
| LegalBench (Vals AI) | Not measured | 80.5% | Only GPT-4 Turbo measured |
| MedQA (Vals AI) | Not measured | 82.0% | Only GPT-4 Turbo measured |
| MMLU Pro (Artificial Analysis run) | 41.7% | 69.4% | GPT-4 Turbo |
| MMLU | Not measured | 81.3% | Only GPT-4 Turbo measured |
| LMArena Text: science | 1306 | 1256 | Gemma 3 4B |
| LMArena Text: medicine and healthcare | 1342 | 1239 | Gemma 3 4B |
| LMArena Text: legal and government | 1327 | 1265 | Gemma 3 4B |
| LMArena Text: business and finance | 1311 | 1222 | Gemma 3 4B |
| Neither model has a result on the other 10 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemma 3 4B | GPT-4 Turbo | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index85
| Benchmark | Gemma 3 4B | GPT-4 Turbo | 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% | 61.6% | GPT-4 Turbo |
| LMCA | 2.8% | 9.8% | GPT-4 Turbo |
| LMArena Text | 1291 | 1272 | Gemma 3 4B |
| LMArena Text: instruction following | 1239 | 1249 | GPT-4 Turbo |
| LMArena Text: creative writing | 1270 | 1268 | Gemma 3 4B |
| LMArena Vision | Not measured | 1090 | Only GPT-4 Turbo measured |
| LMArena Text: English | 1300 | 1291 | Gemma 3 4B |
| LMArena Text: non-English | 1273 | 1245 | Gemma 3 4B |
| LMArena Text: Chinese | Not measured | 1242 | Only GPT-4 Turbo measured |
| LMArena Text: German | 1281 | 1259 | Gemma 3 4B |
| LMArena Text: Japanese | Not measured | 1194 | Only GPT-4 Turbo measured |
| LMArena Text: Korean | Not measured | 1187 | Only GPT-4 Turbo measured |
| LMArena Text: French | Not measured | 1276 | Only GPT-4 Turbo measured |
| LMArena Text: Spanish | Not measured | 1260 | Only GPT-4 Turbo measured |
| LMArena Text: Russian | 1294 | 1259 | Gemma 3 4B |
| LMArena Text: Polish | Not measured | 1302 | Only GPT-4 Turbo measured |
| LMArena Text: multi-turn | 1255 | 1267 | GPT-4 Turbo |
| LMArena Text: longer queries | 1273 | 1254 | Gemma 3 4B |
| LMArena Text: writing and language | 1288 | 1279 | Gemma 3 4B |
| LMArena Text: entertainment and media | 1233 | 1258 | GPT-4 Turbo |
| LMArena Vision: English | Not measured | 1097 | Only GPT-4 Turbo measured |
| LMArena Vision: Chinese | Not measured | 1045 | Only GPT-4 Turbo measured |
| Neither model has a result on the other 22 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemma 3 4B | GPT-4 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 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-4 Turbo |
|---|---|---|
| Maker | OpenAI | |
| Weights | Open weights | No public weights listed |
| Released | March 13, 2025 | April 9, 2024 |
| Context window | 131K | 128K |
| Input $/Mtok | $0.045 | $10.00 |
| Output $/Mtok | $0.075 | $30.00 |
| 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-4 Turbo for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Gemma 3 4B for output tokens 400.0x cheaper ($0.075 vs $30.00 per million) and a larger context window (131K against 128K tokens).
Which is cheaper, Gemma 3 4B or GPT-4 Turbo?
Gemma 3 4B costs $0.045 per million input tokens and $0.075 per million output tokens; GPT-4 Turbo costs $10.00 and $30.00. Cost per solved task is known for neither of them.
Which has the bigger context window, Gemma 3 4B or GPT-4 Turbo?
Gemma 3 4B, with 131K tokens against 128K.
How many benchmarks is the Gemma 3 4B vs GPT-4 Turbo 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-4 Turbo's on 0. Where only one model has a result, the page says so instead of naming a winner.