GPT-3.5 Turbo (older v0613) vs GPT-4 Turbo
GPT-3.5 Turbo (older v0613) costs 15.0x less per output token than GPT-4 Turbo ($2.00 vs $30.00 per million). Updated October 23, 2026; 1 benchmarks have a result for both.
GPT-3.5 Turbo (older v0613) or GPT-4 Turbo: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick GPT-3.5 Turbo (older v0613) for output tokens 15.0x cheaper ($2.00 vs $30.00 per million) and work measured by MMLU, where it leads.
- Shared results
- 1
- GPT-3.5 Turbo (older v0613) only
- 2
- GPT-4 Turbo only
- 41
- Like-for-like categories
- 1 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-4 Turbo has a AA Intelligence Index result, 7.0.
Directional only
Long documents
Listed context window
GPT-4 Turbo
GPT-4 Turbo takes 128K tokens against 4K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
GPT-3.5 Turbo (older v0613)
GPT-3.5 Turbo (older v0613) $0.0020, GPT-4 Turbo $0.025 per turn.
List prices
At a glance
- Coding index
- GPT-3.5 Turbo (older v0613) not benchmarked, GPT-4 Turbo not benchmarked.
- Widest category
- Knowledge, 12.4 points apart on average over 1 shared benchmark.
- Widest single gap
- MMLU: GPT-3.5 Turbo (older v0613) 68.9% against GPT-4 Turbo 81.3%, 12.4 points apart on a 0 to 100 scale.
- Output price
- GPT-3.5 Turbo (older v0613) $2.00 against GPT-4 Turbo $30.00 per million output tokens, 15.0x apart.
- Context window
- GPT-3.5 Turbo (older v0613) takes 4K 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 | GPT-3.5 Turbo (older v0613) | Bars | GPT-4 Turbo | Ties |
|---|---|---|---|---|
| Knowledge | 0 | 1 | ||
| All shared boards | 0 | 1 |
CodingDirectional onlySix of these boards feed the Vibe Coding Index4 boards, none shared
| Benchmark | GPT-3.5 Turbo (older v0613) | GPT-4 Turbo | Leader |
|---|---|---|---|
| AA Coding Index | Not measured | 21.5 | Only GPT-4 Turbo measured |
| LiveCodeBench (Artificial Analysis run) | Not measured | 29.1% | Only GPT-4 Turbo measured |
| LMArena Text: coding | Not measured | 1268 | Only GPT-4 Turbo measured |
| LMArena Text: software and IT | Not measured | 1267 | Only GPT-4 Turbo measured |
| Neither model has a result on the other 39 boards here | |||
IntelligenceDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | GPT-3.5 Turbo (older v0613) | GPT-4 Turbo | Leader |
|---|---|---|---|
| AA Intelligence Index | Not measured | 7.0 | Only GPT-4 Turbo measured |
| Epoch Capabilities Index | Not measured | 127.3 | Only GPT-4 Turbo measured |
Agentic workNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GPT-3.5 Turbo (older v0613) | GPT-4 Turbo | Leader |
|---|---|---|---|
| Neither model has a result on 42 boards here | |||
ReasoningDirectional onlyPublic boards, shown as published, kept out of the index4 boards, none shared
| Benchmark | GPT-3.5 Turbo (older v0613) | GPT-4 Turbo | Leader |
|---|---|---|---|
| Humanity's Last Exam (Artificial Analysis run) | Not measured | 3.1% | Only GPT-4 Turbo measured |
| LMArena Text: hard prompts | Not measured | 1251 | Only GPT-4 Turbo measured |
| LMArena Text: expert | Not measured | 1223 | Only GPT-4 Turbo measured |
| LMArena Text: hard prompts in English | Not measured | 1266 | Only GPT-4 Turbo measured |
| Neither model has a result on the other 19 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index3 boards, none shared
| Benchmark | GPT-3.5 Turbo (older v0613) | GPT-4 Turbo | Leader |
|---|---|---|---|
| GSM8K | 46.9% | Not measured | Only GPT-3.5 Turbo (older v0613) measured |
| LMArena Text: math | Not measured | 1272 | Only GPT-4 Turbo measured |
| LMArena Text: mathematical work | Not measured | 1268 | Only GPT-4 Turbo measured |
| Neither model has a result on the other 11 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | GPT-3.5 Turbo (older v0613) | 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) | Not measured | 69.4% | Only GPT-4 Turbo measured |
| MMLU | 68.9% | 81.3% | GPT-4 Turbo |
| LMArena Text: science | Not measured | 1256 | Only GPT-4 Turbo measured |
| LMArena Text: medicine and healthcare | Not measured | 1239 | Only GPT-4 Turbo measured |
| LMArena Text: legal and government | Not measured | 1265 | Only GPT-4 Turbo measured |
| LMArena Text: business and finance | Not measured | 1222 | Only GPT-4 Turbo measured |
| 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 | GPT-3.5 Turbo (older v0613) | GPT-4 Turbo | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralDirectional onlyPublic boards, shown as published, kept out of the index23 boards, none shared
| Benchmark | GPT-3.5 Turbo (older v0613) | GPT-4 Turbo | Leader |
|---|---|---|---|
| DTBench | Not measured | 61.6% | Only GPT-4 Turbo measured |
| LMCA | Not measured | 9.8% | Only GPT-4 Turbo measured |
| WinoGrande | 81.6% | Not measured | Only GPT-3.5 Turbo (older v0613) measured |
| LMArena Text | Not measured | 1272 | Only GPT-4 Turbo measured |
| LMArena Text: instruction following | Not measured | 1249 | Only GPT-4 Turbo measured |
| LMArena Text: creative writing | Not measured | 1268 | Only GPT-4 Turbo measured |
| LMArena Vision | Not measured | 1090 | Only GPT-4 Turbo measured |
| LMArena Text: English | Not measured | 1291 | Only GPT-4 Turbo measured |
| LMArena Text: non-English | Not measured | 1245 | Only GPT-4 Turbo measured |
| LMArena Text: Chinese | Not measured | 1242 | Only GPT-4 Turbo measured |
| LMArena Text: German | Not measured | 1259 | Only GPT-4 Turbo measured |
| 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 | Not measured | 1259 | Only GPT-4 Turbo measured |
| LMArena Text: Polish | Not measured | 1302 | Only GPT-4 Turbo measured |
| LMArena Text: multi-turn | Not measured | 1267 | Only GPT-4 Turbo measured |
| LMArena Text: longer queries | Not measured | 1253 | Only GPT-4 Turbo measured |
| LMArena Text: writing and language | Not measured | 1279 | Only GPT-4 Turbo measured |
| LMArena Text: entertainment and media | Not measured | 1258 | Only GPT-4 Turbo measured |
| 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 25 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GPT-3.5 Turbo (older v0613) | 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
GPT-3.5 Turbo (older v0613) costs less
Repository review
50K input + 3K output tokens
GPT-3.5 Turbo (older v0613) costs less
Agent loop
200K input + 10K output tokens
GPT-3.5 Turbo (older v0613) 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 | OpenAI GPT-3.5 Turbo (older v0613) | OpenAI GPT-4 Turbo |
|---|---|---|
| Maker | OpenAI | OpenAI |
| Weights | No public weights listed | No public weights listed |
| Released | January 25, 2024 | April 9, 2024 |
| Context window | 4K | 128K |
| Input $/Mtok | $1.00 | $10.00 |
| Output $/Mtok | $2.00 | $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 GPT-3.5 Turbo (older v0613) better than GPT-4 Turbo for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick GPT-3.5 Turbo (older v0613) for output tokens 15.0x cheaper ($2.00 vs $30.00 per million) and work measured by MMLU, where it leads.
Which is cheaper, GPT-3.5 Turbo (older v0613) or GPT-4 Turbo?
GPT-3.5 Turbo (older v0613) costs $1.00 per million input tokens and $2.00 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, GPT-3.5 Turbo (older v0613) or GPT-4 Turbo?
GPT-4 Turbo, with 128K tokens against 4K.
How many benchmarks is the GPT-3.5 Turbo (older v0613) vs GPT-4 Turbo comparison based on?
1 benchmarks have a result for both models as of October 23, 2026. GPT-3.5 Turbo (older v0613)'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.