o4 Mini vs Qwen3.8 27B
Qwen3.8 27B leads o4 Mini on the Artificial Analysis Intelligence Index, 33.7 to 16.7, and Qwen3.8 27B costs 1.7x less per output token ($2.55 vs $4.40 per million). Updated October 21, 2026; 55 benchmarks have a result for both.
o4 Mini or Qwen3.8 27B: which to pick
Only Qwen3.8 27B has a coding score, 74.2 on the Vibe Coding Index; o4 Mini has none, so the two cannot be ranked on coding.
Pick o4 Mini for work measured by TaxEval (Vals AI) or MedCode (Vals AI), where it leads.
- Shared results
- 55
- o4 Mini only
- 43
- Qwen3.8 27B only
- 30
- 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
Only Qwen3.8 27B has a coding score, so no call is made.
One side not measured
Intelligence
AA Intelligence Index
Qwen3.8 27B
Qwen3.8 27B leads, 33.7 against 16.7.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only Qwen3.8 27B has a AA Agentic Index result, 45.8.
Directional only
Reasoning
GPQA Diamond
Qwen3.8 27B
Qwen3.8 27B leads, 88.9% against 74.5%.
Same board, both measured
Math
AIME
No call
Only o4 Mini has a AIME result, 83.7%.
Directional only
Knowledge
MMLU Pro
Qwen3.8 27B
Qwen3.8 27B leads, 84.3% against 80.6%.
Same board, both measured
Web design
Design Arena: websites
No call
Only o4 Mini has a Design Arena: websites result, 985.
Directional only
Long documents
Listed context window
Qwen3.8 27B
Qwen3.8 27B takes 1M tokens against 200K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Qwen3.8 27B
o4 Mini $0.0033, Qwen3.8 27B $0.0017 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only Qwen3.8 27B has the number, $1.26.
Directional only
At a glance
- Coding index
- o4 Mini not benchmarked, Qwen3.8 27B 74.2 from 4 benchmarks.
- Widest category
- Reasoning, 17.7 points apart on average over 11 shared benchmarks.
- Widest single gap
- ARC-AGI-2: o4 Mini 6.1% against Qwen3.8 27B 42.4%, 36.3 points apart on a 0 to 100 scale.
- Output price
- o4 Mini $4.40 against Qwen3.8 27B $2.55 per million output tokens, 1.7x apart.
- Context window
- o4 Mini takes 200K tokens, Qwen3.8 27B 1M.
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 | o4 Mini | Bars | Qwen3.8 27B | Ties |
|---|---|---|---|---|
| Coding | 0 | 3 | ||
| Intelligence | 0 | 2 | ||
| Reasoning | 0 | 11 | ||
| Math | 0 | 2 | ||
| Knowledge | 2 | 7 | ||
| General | 0 | 28 | ||
| All shared boards | 2 | 53 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index03
| Benchmark | o4 Mini | Qwen3.8 27B | Leader |
|---|---|---|---|
| Aider polyglot | 72.0% | Not measured | Only o4 Mini measured |
| SWE-bench Verified | 64.6% | Not measured | Only o4 Mini measured |
| SWE-bench Multimodal | 33.9% | Not measured | Only o4 Mini measured |
| LiveBench coding | Not measured | 86.8% | Only Qwen3.8 27B measured |
| LMArena WebDev | Not measured | 1593 | Only Qwen3.8 27B measured |
| AA Coding Index | Not measured | 68.1 | Only Qwen3.8 27B measured |
| LiveCodeBench (Vals AI) | 82.2% | 84.0% | Qwen3.8 27B |
| SWE-bench Verified (Vals AI) | Not measured | 86.0% | Only Qwen3.8 27B measured |
| Vibe Code Bench (Vals AI) | Not measured | 64.8% | Only Qwen3.8 27B measured |
| Code migration (Vals AI) | Not measured | 14.2% | Only Qwen3.8 27B measured |
| ProgramBench (Vals AI) | Not measured | 0.0% | Only Qwen3.8 27B measured |
| IOI (Vals AI) | Not measured | 39.1% | Only Qwen3.8 27B measured |
| Terminal-Bench 2.1 (Vals AI) | Not measured | 58.4% | Only Qwen3.8 27B measured |
| SkillsBench (Vals AI) | Not measured | 38.1% | Only Qwen3.8 27B measured |
| CyberBench (Vals AI) | Not measured | 67.8% | Only Qwen3.8 27B measured |
| LiveCodeBench (Artificial Analysis run) | 85.9% | Not measured | Only o4 Mini measured |
| SciCode | Not measured | 46.6% | Only Qwen3.8 27B measured |
| Terminal-Bench Hard (Artificial Analysis run) | 15.2% | Not measured | Only o4 Mini measured |
| GSO-Bench | 3.6% | Not measured | Only o4 Mini measured |
| WeirdML | 52.6% | Not measured | Only o4 Mini measured |
| LMArena Text: coding | 1368 | 1483 | Qwen3.8 27B |
| LMArena Text: software and IT | 1367 | 1478 | Qwen3.8 27B |
| LMArena WebDev: React | Not measured | 1602 | Only Qwen3.8 27B measured |
| LMArena WebDev: HTML | Not measured | 1570 | Only Qwen3.8 27B measured |
| LMArena WebDev: from an image | Not measured | 1577 | Only Qwen3.8 27B measured |
| SciCode | 10.8% | Not measured | Only o4 Mini measured |
| Neither model has a result on the other 13 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | o4 Mini | Qwen3.8 27B | Leader |
|---|---|---|---|
| AA Intelligence Index | 16.7 | 33.7 | Qwen3.8 27B |
| Epoch Capabilities Index | 145.5 | 151.7 | Qwen3.8 27B |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index13 boards, none shared
| Benchmark | o4 Mini | Qwen3.8 27B | Leader |
|---|---|---|---|
| AA Agentic Index | Not measured | 45.8 | Only Qwen3.8 27B measured |
| LiveBench research | Not measured | 59.4% | Only Qwen3.8 27B measured |
| LiveBench browser | Not measured | 32.0% | Only Qwen3.8 27B measured |
| LiveBench long agentic | Not measured | 27.9% | Only Qwen3.8 27B measured |
| Finance Agent (Vals AI) | Not measured | 48.5% | Only Qwen3.8 27B measured |
| Tax Agent Bench (Vals AI) | Not measured | 30.9% | Only Qwen3.8 27B measured |
| Legal Research Bench (Vals AI) | Not measured | 36.1% | Only Qwen3.8 27B measured |
| Excel Modeling Benchmark (Vals AI) | Not measured | 59.7% | Only Qwen3.8 27B measured |
| Terminal-Bench-Science (Vals AI) | Not measured | 0.0% | Only Qwen3.8 27B measured |
| τ²-Bench Telecom | 55.6% | Not measured | Only o4 Mini measured |
| APEX-Agents | Not measured | 47.5% | Only Qwen3.8 27B measured |
| METR time horizons | 63.9% | Not measured | Only o4 Mini measured |
| Berkeley Function Calling Leaderboard (BFCL) | 53.2% | Not measured | Only o4 Mini measured |
| Neither model has a result on the other 28 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index011
| Benchmark | o4 Mini | Qwen3.8 27B | Leader |
|---|---|---|---|
| GPQA Diamond | 74.5% | 88.9% | Qwen3.8 27B |
| LiveBench reasoning | Not measured | 80.0% | Only Qwen3.8 27B measured |
| MMMU Pro (Vals AI) | 79.7% | 83.9% | Qwen3.8 27B |
| GPQA Diamond (Artificial Analysis run) | 78.4% | 90.5% | Qwen3.8 27B |
| Humanity's Last Exam (Artificial Analysis run) | 16.5% | 33.9% | Qwen3.8 27B |
| GPQA Diamond (Epoch AI run) | 79.6% | Not measured | Only o4 Mini measured |
| ARC-AGI | 58.7% | 87.5% | Qwen3.8 27B |
| ARC-AGI-2 | 6.1% | 42.4% | Qwen3.8 27B |
| SimpleBench | 38.7% | Not measured | Only o4 Mini measured |
| Chess puzzles | 22.0% | Not measured | Only o4 Mini measured |
| Mystery game puzzles | 5.0% | Not measured | Only o4 Mini measured |
| VPCT | 57.5% | Not measured | Only o4 Mini measured |
| LMArena Text: hard prompts | 1351 | 1459 | Qwen3.8 27B |
| LMArena Text: expert | 1343 | 1479 | Qwen3.8 27B |
| LMArena Text: hard prompts in English | 1366 | 1461 | Qwen3.8 27B |
| LMArena Vision: diagrams | 1197 | 1285 | Qwen3.8 27B |
| LMArena Vision: homework | 1234 | 1295 | Qwen3.8 27B |
| Kagi LLM Benchmark | 67.6% | Not measured | Only o4 Mini measured |
| Neither model has a result on the other 4 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | o4 Mini | Qwen3.8 27B | Leader |
|---|---|---|---|
| AIME | 83.7% | Not measured | Only o4 Mini measured |
| MATH 500 (Vals AI) | 94.2% | Not measured | Only o4 Mini measured |
| MGSM (Vals AI) | 93.4% | Not measured | Only o4 Mini measured |
| AA Math Index | 90.7 | Not measured | Only o4 Mini measured |
| AIME 2025 (Artificial Analysis run) | 90.7% | Not measured | Only o4 Mini measured |
| OTIS Mock AIME | 81.7% | Not measured | Only o4 Mini measured |
| FrontierMath, tiers 1 to 3 | 36.1% | Not measured | Only o4 Mini measured |
| FrontierMath, tier 4 | 4.9% | Not measured | Only o4 Mini measured |
| ProofBench | Not measured | 16.0% | Only Qwen3.8 27B measured |
| MATH level 5 | 97.8% | Not measured | Only o4 Mini measured |
| LMArena Text: math | 1389 | 1455 | Qwen3.8 27B |
| LMArena Text: mathematical work | 1379 | 1458 | Qwen3.8 27B |
| Neither model has a result on the other 1 board here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index27
| Benchmark | o4 Mini | Qwen3.8 27B | Leader |
|---|---|---|---|
| MMLU Pro | 80.6% | 84.3% | Qwen3.8 27B |
| LegalBench (Vals AI) | 79.2% | 82.4% | Qwen3.8 27B |
| CorpFin (Vals AI) | 59.0% | Not measured | Only o4 Mini measured |
| TaxEval (Vals AI) | 74.8% | 70.8% | o4 Mini |
| MortgageTax (Vals AI) | 64.8% | 64.9% | Qwen3.8 27B |
| MedQA (Vals AI) | 96.0% | Not measured | Only o4 Mini measured |
| MedCode (Vals AI) | 33.8% | 28.7% | o4 Mini |
| MMLU Pro (Artificial Analysis run) | 83.2% | Not measured | Only o4 Mini measured |
| SimpleQA Verified | 19.6% | Not measured | Only o4 Mini measured |
| GeoBench | 64.0% | Not measured | Only o4 Mini measured |
| LMArena Text: science | 1354 | 1467 | Qwen3.8 27B |
| LMArena Text: medicine and healthcare | 1361 | 1437 | Qwen3.8 27B |
| LMArena Text: legal and government | 1350 | 1447 | Qwen3.8 27B |
| LMArena Text: business and finance | 1325 | 1438 | Qwen3.8 27B |
| FACTS Benchmark Suite | 28.1% | Not measured | Only o4 Mini measured |
| Neither model has a result on the other 3 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index7 boards, none shared
| Benchmark | o4 Mini | Qwen3.8 27B | Leader |
|---|---|---|---|
| Design Arena: websites | 985 | Not measured | Only o4 Mini measured |
| Design Arena: UI components | 983 | Not measured | Only o4 Mini measured |
| Design Arena: game dev | 1003 | Not measured | Only o4 Mini measured |
| Design Arena: data viz | 990 | Not measured | Only o4 Mini measured |
| Design Arena: 3D | 854 | Not measured | Only o4 Mini measured |
| Design Arena: all code categories | 976 | Not measured | Only o4 Mini measured |
| LiveBench design | Not measured | 75.9% | Only Qwen3.8 27B measured |
| Neither model has a result on the other 2 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index028
| Benchmark | o4 Mini | Qwen3.8 27B | Leader |
|---|---|---|---|
| LiveBench data analysis | Not measured | 76.6% | Only Qwen3.8 27B measured |
| SAGE (Vals AI) | 41.1% | 52.4% | Qwen3.8 27B |
| MedScribe (Vals AI) | 69.1% | 83.8% | Qwen3.8 27B |
| IFBench | 68.7% | Not measured | Only o4 Mini measured |
| Long context reasoning (AA-LCR) | 61.0% | 82.0% | Qwen3.8 27B |
| DTBench | 77.6% | 88.0% | Qwen3.8 27B |
| LMCA | 26.5% | 41.4% | Qwen3.8 27B |
| Fiction.LiveBench | 77.8% | Not measured | Only o4 Mini measured |
| Creative writing (Lech Mazur) | 75.0% | Not measured | Only o4 Mini measured |
| Surface Evolver Bench | Not measured | 45.0% | Only Qwen3.8 27B measured |
| LMArena Text | 1353 | 1440 | Qwen3.8 27B |
| LMArena Text: instruction following | 1320 | 1436 | Qwen3.8 27B |
| LMArena Text: creative writing | 1294 | 1383 | Qwen3.8 27B |
| LMArena Vision | 1194 | 1269 | Qwen3.8 27B |
| LMArena Text: English | 1367 | 1448 | Qwen3.8 27B |
| LMArena Text: non-English | 1337 | 1428 | Qwen3.8 27B |
| LMArena Text: Chinese | 1354 | 1501 | Qwen3.8 27B |
| LMArena Text: German | 1336 | 1443 | Qwen3.8 27B |
| LMArena Text: Japanese | 1308 | 1387 | Qwen3.8 27B |
| LMArena Text: Korean | 1312 | 1384 | Qwen3.8 27B |
| LMArena Text: French | 1364 | 1465 | Qwen3.8 27B |
| LMArena Text: Spanish | 1346 | 1449 | Qwen3.8 27B |
| LMArena Text: Russian | 1334 | 1415 | Qwen3.8 27B |
| LMArena Text: Polish | 1358 | 1439 | Qwen3.8 27B |
| LMArena Text: multi-turn | 1350 | 1440 | Qwen3.8 27B |
| LMArena Text: longer queries | 1315 | 1448 | Qwen3.8 27B |
| LMArena Text: writing and language | 1316 | 1405 | Qwen3.8 27B |
| LMArena Text: entertainment and media | 1308 | 1393 | Qwen3.8 27B |
| LMArena Vision: English | 1200 | 1267 | Qwen3.8 27B |
| LMArena Vision: Chinese | 1218 | 1326 | Qwen3.8 27B |
| LMArena Vision: reading text | 1197 | 1281 | Qwen3.8 27B |
| LMArena Vision: humor | 1199 | 1235 | Qwen3.8 27B |
| LMArena Vision: captioning | 1204 | Not measured | Only o4 Mini measured |
| LMArena Vision: recognizing things | 1218 | Not measured | Only o4 Mini measured |
| LMArena Vision: creative writing | 1166 | 1237 | Qwen3.8 27B |
| Neither model has a result on the other 11 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | o4 Mini | Qwen3.8 27B | 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
Qwen3.8 27B costs less
Repository review
50K input + 3K output tokens
Qwen3.8 27B costs less
Agent loop
200K input + 10K output tokens
Qwen3.8 27B 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 o4 Mini | Qwen3.8 27B |
|---|---|---|
| Maker | OpenAI | Qwen |
| Weights | No public weights listed | Open weights |
| Released | April 16, 2025 | August 14, 2026 |
| Context window | 200K | 1M |
| Input $/Mtok | $1.10 | $0.425 |
| Output $/Mtok | $4.40 | $2.55 |
| Coding index | Not benchmarked | 74.2 |
| Independent measurements behind it | 0 | 4 |
| Cost per solved task | Not computable | $1.26 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is o4 Mini better than Qwen3.8 27B for coding?
Only Qwen3.8 27B has a coding score, 74.2 on the Vibe Coding Index; o4 Mini has none, so the two cannot be ranked on coding. Pick o4 Mini for work measured by TaxEval (Vals AI) or MedCode (Vals AI), where it leads.
Which is cheaper, o4 Mini or Qwen3.8 27B?
o4 Mini costs $1.10 per million input tokens and $4.40 per million output tokens; Qwen3.8 27B costs $0.425 and $2.55. Cost per solved task is known for Qwen3.8 27B only.
Which has the bigger context window, o4 Mini or Qwen3.8 27B?
Qwen3.8 27B, with 1M tokens against 200K.
How many benchmarks is the o4 Mini vs Qwen3.8 27B comparison based on?
55 benchmarks have a result for both models as of October 21, 2026. o4 Mini's coding index rests on 0 independent benchmarks and Qwen3.8 27B's on 4. Where only one model has a result, the page says so instead of naming a winner.