Gemma 2 27B vs GPT-3.5 Turbo
Gemma 2 27B costs 2.3x less per output token than GPT-3.5 Turbo ($0.65 vs $1.50 per million). Updated October 22, 2026; 4 benchmarks have a result for both.
Gemma 2 27B or GPT-3.5 Turbo: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Gemma 2 27B for output tokens 2.3x cheaper ($0.65 vs $1.50 per million) and work measured by DTBench or LMCA, where it leads.
- Shared results
- 4
- Gemma 2 27B only
- 32
- GPT-3.5 Turbo only
- 9
- Like-for-like categories
- 2 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 GPT-3.5 Turbo has a AA Intelligence Index result, 5.5.
Directional only
Reasoning
GPQA Diamond
No call
Only GPT-3.5 Turbo has a GPQA Diamond result, 30.6%.
Directional only
Long documents
Listed context window
GPT-3.5 Turbo
GPT-3.5 Turbo takes 16K tokens against 8K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Gemma 2 27B
Gemma 2 27B $0.0010, GPT-3.5 Turbo $0.0013 per turn.
List prices
At a glance
- Coding index
- Gemma 2 27B not benchmarked, GPT-3.5 Turbo not benchmarked.
- Widest category
- Knowledge, 5.4 points apart on average over 2 shared benchmarks.
- Widest single gap
- MMLU: Gemma 2 27B 75.7% against GPT-3.5 Turbo 68.5%, 7.2 points apart on a 0 to 100 scale.
- Output price
- Gemma 2 27B $0.65 against GPT-3.5 Turbo $1.50 per million output tokens, 2.3x apart.
- Context window
- Gemma 2 27B takes 8K tokens, GPT-3.5 Turbo 16K.
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 2 27B | Bars | GPT-3.5 Turbo | Ties |
|---|---|---|---|---|
| Knowledge | 2 | 0 | ||
| General | 0 | 2 | ||
| All shared boards | 2 | 2 |
CodingDirectional onlySix of these boards feed the Vibe Coding Index4 boards, none shared
| Benchmark | Gemma 2 27B | GPT-3.5 Turbo | Leader |
|---|---|---|---|
| AA Coding Index | Not measured | 10.7 | Only GPT-3.5 Turbo measured |
| LMArena Text: coding | 1211 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: software and IT | 1228 | Not measured | Only Gemma 2 27B measured |
| SciCode | Not measured | 0.0% | Only GPT-3.5 Turbo measured |
| Neither model has a result on the other 35 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Gemma 2 27B | GPT-3.5 Turbo | Leader |
|---|---|---|---|
| AA Intelligence Index | Not measured | 5.5 | Only GPT-3.5 Turbo measured |
| Epoch Capabilities Index | 122.1 | Not measured | Only Gemma 2 27B measured |
Agentic workNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemma 2 27B | GPT-3.5 Turbo | Leader |
|---|---|---|---|
| Neither model has a result on 41 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
| Benchmark | Gemma 2 27B | GPT-3.5 Turbo | Leader |
|---|---|---|---|
| GPQA Diamond | Not measured | 30.6% | Only GPT-3.5 Turbo measured |
| GPQA Diamond (Artificial Analysis run) | Not measured | 29.7% | Only GPT-3.5 Turbo measured |
| GPQA Diamond (Epoch AI run) | 36.5% | Not measured | Only Gemma 2 27B measured |
| BIG-Bench Hard | Not measured | 61.6% | Only GPT-3.5 Turbo measured |
| LMArena Text: hard prompts | 1198 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: expert | 1171 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: hard prompts in English | 1206 | Not measured | Only Gemma 2 27B measured |
| Neither model has a result on the other 15 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index5 boards, none shared
| Benchmark | Gemma 2 27B | GPT-3.5 Turbo | Leader |
|---|---|---|---|
| OTIS Mock AIME | 1.4% | Not measured | Only Gemma 2 27B measured |
| MATH level 5 | 27.9% | Not measured | Only Gemma 2 27B measured |
| GSM8K | Not measured | 57.8% | Only GPT-3.5 Turbo measured |
| LMArena Text: math | 1212 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: mathematical work | 1221 | Not measured | Only Gemma 2 27B measured |
| Neither model has a result on the other 8 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Gemma 2 27B | GPT-3.5 Turbo | Leader |
|---|---|---|---|
| LegalBench (Vals AI) | 68.0% | 64.4% | Gemma 2 27B |
| MedQA (Vals AI) | Not measured | 58.5% | Only GPT-3.5 Turbo measured |
| MMLU Pro (Artificial Analysis run) | Not measured | 46.2% | Only GPT-3.5 Turbo measured |
| MMLU | 75.7% | 68.5% | Gemma 2 27B |
| LMArena Text: science | 1228 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: medicine and healthcare | 1198 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: legal and government | 1241 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: business and finance | 1208 | Not measured | Only Gemma 2 27B measured |
| 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 2 27B | GPT-3.5 Turbo | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Gemma 2 27B | GPT-3.5 Turbo | Leader |
|---|---|---|---|
| DTBench | 48.0% | 48.5% | GPT-3.5 Turbo |
| LMCA | 7.1% | 9.7% | GPT-3.5 Turbo |
| LMArena Text | 1231 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: instruction following | 1206 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: creative writing | 1240 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: English | 1242 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: non-English | 1217 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: Chinese | 1221 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: German | 1209 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: Japanese | 1175 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: Korean | 1174 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: French | 1248 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: Spanish | 1228 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: Russian | 1234 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: Polish | 1277 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: multi-turn | 1224 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: longer queries | 1230 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: writing and language | 1249 | Not measured | Only Gemma 2 27B measured |
| LMArena Text: entertainment and media | 1212 | Not measured | Only Gemma 2 27B measured |
| Neither model has a result on the other 27 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemma 2 27B | GPT-3.5 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 2 27B costs less
Repository review
50K input + 3K output tokens
GPT-3.5 Turbo costs less
Agent loop
200K input + 10K output tokens
GPT-3.5 Turbo 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 2 27B | OpenAI GPT-3.5 Turbo |
|---|---|---|
| Maker | OpenAI | |
| Weights | Open weights | No public weights listed |
| Released | July 13, 2024 | May 28, 2023 |
| Context window | 8K | 16K |
| Input $/Mtok | $0.65 | $0.50 |
| Output $/Mtok | $0.65 | $1.50 |
| 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 2 27B better than GPT-3.5 Turbo for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Gemma 2 27B for output tokens 2.3x cheaper ($0.65 vs $1.50 per million) and work measured by DTBench or LMCA, where it leads.
Which is cheaper, Gemma 2 27B or GPT-3.5 Turbo?
Gemma 2 27B costs $0.65 per million input tokens and $0.65 per million output tokens; GPT-3.5 Turbo costs $0.50 and $1.50. Cost per solved task is known for neither of them.
Which has the bigger context window, Gemma 2 27B or GPT-3.5 Turbo?
GPT-3.5 Turbo, with 16K tokens against 8K.
How many benchmarks is the Gemma 2 27B vs GPT-3.5 Turbo comparison based on?
4 benchmarks have a result for both models as of October 22, 2026. Gemma 2 27B's coding index rests on 0 independent benchmarks and GPT-3.5 Turbo's on 0. Where only one model has a result, the page says so instead of naming a winner.