GPT-4 Turbo vs Qwen2.5 Coder 32B Instruct
GPT-4 Turbo leads Qwen2.5 Coder 32B Instruct on the Artificial Analysis Intelligence Index, 7.0 to 6.7, and Qwen2.5 Coder 32B Instruct costs 30.0x less per output token ($1.00 vs $30.00 per million). Updated October 21, 2026; 28 benchmarks have a result for both.
GPT-4 Turbo or Qwen2.5 Coder 32B Instruct: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick GPT-4 Turbo for a larger context window (128K against 33K tokens) and work measured by LiveCodeBench (Artificial Analysis run) or LMArena Text: coding, where it leads.
- Shared results
- 28
- GPT-4 Turbo only
- 14
- Qwen2.5 Coder 32B Instruct only
- 7
- 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 6.7.
Same board, both measured
Long documents
Listed context window
GPT-4 Turbo
GPT-4 Turbo takes 128K tokens against 33K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen2.5 Coder 32B Instruct
GPT-4 Turbo $0.025, Qwen2.5 Coder 32B Instruct $0.0012 per turn.
List prices
At a glance
- Coding index
- GPT-4 Turbo not benchmarked, Qwen2.5 Coder 32B Instruct not benchmarked.
- Widest category
- General, 8.2 points apart on average over 11 shared benchmarks.
- Widest single gap
- LMArena Text: entertainment and media: GPT-4 Turbo 1258 against Qwen2.5 Coder 32B Instruct 1150, 18.0 points apart on a 0 to 100 scale.
- Output price
- GPT-4 Turbo $30.00 against Qwen2.5 Coder 32B Instruct $1.00 per million output tokens, 30.0x apart.
- Context window
- GPT-4 Turbo takes 128K tokens, Qwen2.5 Coder 32B Instruct 33K.
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-4 Turbo | Bars | Qwen2.5 Coder 32B Instruct | Ties |
|---|---|---|---|---|
| Coding | 1 | 2 | ||
| Intelligence | 2 | 0 | ||
| Reasoning | 1 | 3 | ||
| Math | 2 | 0 | ||
| Knowledge | 5 | 1 | ||
| General | 11 | 0 | ||
| All shared boards | 22 | 6 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index12
| Benchmark | GPT-4 Turbo | Qwen2.5 Coder 32B Instruct | Leader |
|---|---|---|---|
| Aider polyglot | Not measured | 16.4% | Only Qwen2.5 Coder 32B Instruct measured |
| SWE-bench Verified | Not measured | 47.0% | Only Qwen2.5 Coder 32B Instruct measured |
| AA Coding Index | 21.5 | Not measured | Only GPT-4 Turbo measured |
| LiveCodeBench (Artificial Analysis run) | 29.1% | 29.5% | Qwen2.5 Coder 32B Instruct |
| LMArena Text: coding | 1268 | 1276 | Qwen2.5 Coder 32B Instruct |
| LMArena Text: software and IT | 1267 | 1261 | GPT-4 Turbo |
| Neither model has a result on the other 33 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | GPT-4 Turbo | Qwen2.5 Coder 32B Instruct | Leader |
|---|---|---|---|
| AA Intelligence Index | 7.0 | 6.7 | GPT-4 Turbo |
| Epoch Capabilities Index | 127.3 | 119.5 | GPT-4 Turbo |
Agentic workNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GPT-4 Turbo | Qwen2.5 Coder 32B Instruct | Leader |
|---|---|---|---|
| Neither model has a result on 41 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index13
| Benchmark | GPT-4 Turbo | Qwen2.5 Coder 32B Instruct | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | Not measured | 41.7% | Only Qwen2.5 Coder 32B Instruct measured |
| Humanity's Last Exam (Artificial Analysis run) | 3.1% | 3.5% | Qwen2.5 Coder 32B Instruct |
| LMArena Text: hard prompts | 1251 | 1251 | Qwen2.5 Coder 32B Instruct |
| LMArena Text: expert | 1223 | 1221 | GPT-4 Turbo |
| LMArena Text: hard prompts in English | 1266 | 1269 | Qwen2.5 Coder 32B Instruct |
| Neither model has a result on the other 17 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | GPT-4 Turbo | Qwen2.5 Coder 32B Instruct | Leader |
|---|---|---|---|
| GSM8K | Not measured | 93.0% | Only Qwen2.5 Coder 32B Instruct measured |
| LMArena Text: math | 1272 | 1251 | GPT-4 Turbo |
| LMArena Text: mathematical work | 1268 | 1243 | GPT-4 Turbo |
| Neither model has a result on the other 10 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index51
| Benchmark | GPT-4 Turbo | Qwen2.5 Coder 32B Instruct | Leader |
|---|---|---|---|
| LegalBench (Vals AI) | 80.5% | Not measured | Only GPT-4 Turbo measured |
| MedQA (Vals AI) | 82.0% | Not measured | Only GPT-4 Turbo measured |
| MMLU Pro (Artificial Analysis run) | 69.4% | 63.5% | GPT-4 Turbo |
| MMLU | 81.3% | 79.1% | GPT-4 Turbo |
| AI2 Reasoning Challenge | Not measured | 70.5% | Only Qwen2.5 Coder 32B Instruct measured |
| LMArena Text: science | 1256 | 1221 | GPT-4 Turbo |
| LMArena Text: medicine and healthcare | 1239 | 1187 | GPT-4 Turbo |
| LMArena Text: legal and government | 1265 | 1257 | GPT-4 Turbo |
| LMArena Text: business and finance | 1222 | 1224 | Qwen2.5 Coder 32B Instruct |
| Neither model has a result on the other 9 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GPT-4 Turbo | Qwen2.5 Coder 32B Instruct | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index110
| Benchmark | GPT-4 Turbo | Qwen2.5 Coder 32B Instruct | Leader |
|---|---|---|---|
| DTBench | 61.6% | Not measured | Only GPT-4 Turbo measured |
| LMCA | 9.8% | Not measured | Only GPT-4 Turbo measured |
| HellaSwag | Not measured | 83.0% | Only Qwen2.5 Coder 32B Instruct measured |
| WinoGrande | Not measured | 80.8% | Only Qwen2.5 Coder 32B Instruct measured |
| LMArena Text | 1272 | 1230 | GPT-4 Turbo |
| LMArena Text: instruction following | 1249 | 1223 | GPT-4 Turbo |
| LMArena Text: creative writing | 1268 | 1174 | GPT-4 Turbo |
| LMArena Vision | 1090 | Not measured | Only GPT-4 Turbo measured |
| LMArena Text: English | 1291 | 1248 | GPT-4 Turbo |
| LMArena Text: non-English | 1245 | 1205 | GPT-4 Turbo |
| LMArena Text: Chinese | 1242 | 1222 | GPT-4 Turbo |
| LMArena Text: German | 1259 | Not measured | Only GPT-4 Turbo measured |
| LMArena Text: Japanese | 1194 | Not measured | Only GPT-4 Turbo measured |
| LMArena Text: Korean | 1187 | Not measured | Only GPT-4 Turbo measured |
| LMArena Text: French | 1276 | Not measured | Only GPT-4 Turbo measured |
| LMArena Text: Spanish | 1260 | Not measured | Only GPT-4 Turbo measured |
| LMArena Text: Russian | 1259 | 1228 | GPT-4 Turbo |
| LMArena Text: Polish | 1302 | Not measured | Only GPT-4 Turbo measured |
| LMArena Text: multi-turn | 1267 | 1222 | GPT-4 Turbo |
| LMArena Text: longer queries | 1254 | 1251 | GPT-4 Turbo |
| LMArena Text: writing and language | 1279 | 1191 | GPT-4 Turbo |
| LMArena Text: entertainment and media | 1258 | 1150 | GPT-4 Turbo |
| LMArena Vision: English | 1097 | Not measured | Only GPT-4 Turbo measured |
| LMArena Vision: Chinese | 1045 | Not measured | 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 | GPT-4 Turbo | Qwen2.5 Coder 32B Instruct | 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
Qwen2.5 Coder 32B Instruct costs less
Repository review
50K input + 3K output tokens
Qwen2.5 Coder 32B Instruct costs less
Agent loop
200K input + 10K output tokens
Qwen2.5 Coder 32B Instruct 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-4 Turbo | Qwen2.5 Coder 32B Instruct |
|---|---|---|
| Maker | OpenAI | Qwen |
| Weights | No public weights listed | Open weights |
| Released | April 9, 2024 | November 11, 2024 |
| Context window | 128K | 33K |
| Input $/Mtok | $10.00 | $0.66 |
| Output $/Mtok | $30.00 | $1.00 |
| 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-4 Turbo better than Qwen2.5 Coder 32B Instruct for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick GPT-4 Turbo for a larger context window (128K against 33K tokens) and work measured by LiveCodeBench (Artificial Analysis run) or LMArena Text: coding, where it leads.
Which is cheaper, GPT-4 Turbo or Qwen2.5 Coder 32B Instruct?
GPT-4 Turbo costs $10.00 per million input tokens and $30.00 per million output tokens; Qwen2.5 Coder 32B Instruct costs $0.66 and $1.00. Cost per solved task is known for neither of them.
Which has the bigger context window, GPT-4 Turbo or Qwen2.5 Coder 32B Instruct?
GPT-4 Turbo, with 128K tokens against 33K.
How many benchmarks is the GPT-4 Turbo vs Qwen2.5 Coder 32B Instruct comparison based on?
28 benchmarks have a result for both models as of October 21, 2026. GPT-4 Turbo's coding index rests on 0 independent benchmarks and Qwen2.5 Coder 32B Instruct's on 0. Where only one model has a result, the page says so instead of naming a winner.