Phi 4 vs Qwen3 VL 235B A22B Instruct
Qwen3 VL 235B A22B Instruct leads Phi 4 on the Artificial Analysis Intelligence Index, 9.9 to 5.9, and Phi 4 costs 13.6x less per output token ($0.14 vs $1.90 per million). Updated October 21, 2026; 38 benchmarks have a result for both.
Phi 4 or Qwen3 VL 235B A22B Instruct: which to pick
Neither model has a coding score yet, so the two cannot be ranked on coding.
Pick Phi 4 for output tokens 13.6x cheaper ($0.14 vs $1.90 per million) and work measured by LiveCodeBench (Artificial Analysis run) or Terminal-Bench Hard (Artificial Analysis run), where it leads.
- Shared results
- 38
- Phi 4 only
- 8
- Qwen3 VL 235B A22B Instruct only
- 11
- Like-for-like categories
- 7 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
Qwen3 VL 235B A22B Instruct
Qwen3 VL 235B A22B Instruct leads, 9.9 against 5.9.
Same board, both measured
Long documents
Listed context window
Qwen3 VL 235B A22B Instruct
Qwen3 VL 235B A22B Instruct takes 262K tokens against 16K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Phi 4
Phi 4 $0.0001, Qwen3 VL 235B A22B Instruct $0.0012 per turn.
List prices
At a glance
- Coding index
- Phi 4 not benchmarked, Qwen3 VL 235B A22B Instruct not benchmarked.
- Widest category
- Agentic work, 54.1 points apart on average over 1 shared benchmark.
- Widest single gap
- AIME 2025 (Artificial Analysis run): Phi 4 18.0% against Qwen3 VL 235B A22B Instruct 88.3%, 70.3 points apart on a 0 to 100 scale.
- Output price
- Phi 4 $0.14 against Qwen3 VL 235B A22B Instruct $1.90 per million output tokens, 13.6x apart.
- Context window
- Phi 4 takes 16K tokens, Qwen3 VL 235B A22B Instruct 262K.
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 | Phi 4 | Bars | Qwen3 VL 235B A22B Instruct | Ties |
|---|---|---|---|---|
| Coding | 0 | 4 | ||
| Intelligence | 0 | 1 | ||
| Agentic work | 0 | 1 | ||
| Reasoning | 0 | 5 | ||
| Math | 0 | 4 | ||
| Knowledge | 0 | 5 | ||
| General | 0 | 18 | ||
| All shared boards | 0 | 38 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index04
| Benchmark | Phi 4 | Qwen3 VL 235B A22B Instruct | Leader |
|---|---|---|---|
| LiveCodeBench (Artificial Analysis run) | 23.1% | 64.6% | Qwen3 VL 235B A22B Instruct |
| Terminal-Bench Hard (Artificial Analysis run) | 3.8% | 11.4% | Qwen3 VL 235B A22B Instruct |
| LMArena Text: coding | 1231 | 1438 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: software and IT | 1228 | 1442 | Qwen3 VL 235B A22B Instruct |
| Neither model has a result on the other 35 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Phi 4 | Qwen3 VL 235B A22B Instruct | Leader |
|---|---|---|---|
| AA Intelligence Index | 5.9 | 9.9 | Qwen3 VL 235B A22B Instruct |
| Epoch Capabilities Index | 130.5 | Not measured | Only Phi 4 measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | Phi 4 | Qwen3 VL 235B A22B Instruct | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 0.0% | 54.1% | Qwen3 VL 235B A22B Instruct |
| BALROG | 11.6% | Not measured | Only Phi 4 measured |
| Berkeley Function Calling Leaderboard (BFCL) | 28.8% | Not measured | Only Phi 4 measured |
| Neither model has a result on the other 38 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index05
| Benchmark | Phi 4 | Qwen3 VL 235B A22B Instruct | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 57.5% | 77.2% | Qwen3 VL 235B A22B Instruct |
| Humanity's Last Exam (Artificial Analysis run) | 3.8% | 11.9% | Qwen3 VL 235B A22B Instruct |
| GPQA Diamond (Epoch AI run) | 56.1% | Not measured | Only Phi 4 measured |
| LMArena Text: hard prompts | 1220 | 1428 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: expert | 1203 | 1442 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: hard prompts in English | 1230 | 1439 | Qwen3 VL 235B A22B Instruct |
| LMArena Vision: diagrams | Not measured | 1255 | Only Qwen3 VL 235B A22B Instruct measured |
| LMArena Vision: homework | Not measured | 1277 | Only Qwen3 VL 235B A22B Instruct measured |
| Neither model has a result on the other 14 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index04
| Benchmark | Phi 4 | Qwen3 VL 235B A22B Instruct | Leader |
|---|---|---|---|
| AA Math Index | 18.0 | 88.3 | Qwen3 VL 235B A22B Instruct |
| AIME 2025 (Artificial Analysis run) | 18.0% | 88.3% | Qwen3 VL 235B A22B Instruct |
| OTIS Mock AIME | 13.8% | Not measured | Only Phi 4 measured |
| MATH level 5 | 64.9% | Not measured | Only Phi 4 measured |
| LMArena Text: math | 1246 | 1425 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: mathematical work | 1254 | 1419 | Qwen3 VL 235B A22B Instruct |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index05
| Benchmark | Phi 4 | Qwen3 VL 235B A22B Instruct | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | 71.4% | 83.6% | Qwen3 VL 235B A22B Instruct |
| MMLU | 84.8% | Not measured | Only Phi 4 measured |
| LMArena Text: science | 1206 | 1444 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: medicine and healthcare | 1202 | 1431 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: legal and government | 1241 | 1438 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: business and finance | 1199 | 1433 | Qwen3 VL 235B A22B Instruct |
| 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 | Phi 4 | Qwen3 VL 235B A22B Instruct | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index018
| Benchmark | Phi 4 | Qwen3 VL 235B A22B Instruct | Leader |
|---|---|---|---|
| IFBench | 23.5% | 56.5% | Qwen3 VL 235B A22B Instruct |
| Long context reasoning (AA-LCR) | 0.0% | 63.7% | Qwen3 VL 235B A22B Instruct |
| Creative writing (Lech Mazur) | 62.6% | Not measured | Only Phi 4 measured |
| LMArena Text | 1217 | 1420 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: instruction following | 1201 | 1406 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: creative writing | 1181 | 1366 | Qwen3 VL 235B A22B Instruct |
| LMArena Vision | Not measured | 1247 | Only Qwen3 VL 235B A22B Instruct measured |
| LMArena Text: English | 1231 | 1429 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: non-English | 1197 | 1405 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: Chinese | 1212 | 1463 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: German | 1222 | 1424 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: Japanese | 1158 | 1385 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: Korean | 1151 | 1394 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: French | 1224 | 1452 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: Spanish | 1234 | 1428 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: Russian | 1209 | 1408 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: Polish | Not measured | 1419 | Only Qwen3 VL 235B A22B Instruct measured |
| LMArena Text: multi-turn | 1206 | 1428 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: longer queries | 1217 | 1419 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: writing and language | 1198 | 1392 | Qwen3 VL 235B A22B Instruct |
| LMArena Text: entertainment and media | 1162 | 1368 | Qwen3 VL 235B A22B Instruct |
| LMArena Vision: English | Not measured | 1256 | Only Qwen3 VL 235B A22B Instruct measured |
| LMArena Vision: Chinese | Not measured | 1283 | Only Qwen3 VL 235B A22B Instruct measured |
| LMArena Vision: reading text | Not measured | 1253 | Only Qwen3 VL 235B A22B Instruct measured |
| LMArena Vision: humor | Not measured | 1209 | Only Qwen3 VL 235B A22B Instruct measured |
| LMArena Vision: captioning | Not measured | 1281 | Only Qwen3 VL 235B A22B Instruct measured |
| LMArena Vision: recognizing things | Not measured | 1249 | Only Qwen3 VL 235B A22B Instruct measured |
| LMArena Vision: creative writing | Not measured | 1220 | Only Qwen3 VL 235B A22B Instruct measured |
| Neither model has a result on the other 18 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Phi 4 | Qwen3 VL 235B A22B 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
Phi 4 costs less
Repository review
50K input + 3K output tokens
Phi 4 costs less
Agent loop
200K input + 10K output tokens
Phi 4 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 | Microsoft Phi 4 | Qwen3 VL 235B A22B Instruct |
|---|---|---|
| Maker | Microsoft | Qwen |
| Weights | Open weights | Open weights |
| Released | January 10, 2025 | September 23, 2025 |
| Context window | 16K | 262K |
| Input $/Mtok | $0.07 | $0.21 |
| Output $/Mtok | $0.14 | $1.90 |
| 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 Phi 4 better than Qwen3 VL 235B A22B Instruct for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Pick Phi 4 for output tokens 13.6x cheaper ($0.14 vs $1.90 per million) and work measured by LiveCodeBench (Artificial Analysis run) or Terminal-Bench Hard (Artificial Analysis run), where it leads.
Which is cheaper, Phi 4 or Qwen3 VL 235B A22B Instruct?
Phi 4 costs $0.07 per million input tokens and $0.14 per million output tokens; Qwen3 VL 235B A22B Instruct costs $0.21 and $1.90. Cost per solved task is known for neither of them.
Which has the bigger context window, Phi 4 or Qwen3 VL 235B A22B Instruct?
Qwen3 VL 235B A22B Instruct, with 262K tokens against 16K.
How many benchmarks is the Phi 4 vs Qwen3 VL 235B A22B Instruct comparison based on?
38 benchmarks have a result for both models as of October 21, 2026. Phi 4's coding index rests on 0 independent benchmarks and Qwen3 VL 235B A22B Instruct's on 0. Where only one model has a result, the page says so instead of naming a winner.