o4 Mini vs Step 3.7 Flash
Updated October 21, 2026. 13 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Neither model has a coding score yet, so the two cannot be ranked on coding.
o4 Mini has the edge where what matters is Terminal-Bench Hard (Artificial Analysis run) or AA Intelligence Index, where it leads.
- Shared results
- 13
- o4 Mini only
- 81
- Step 3.7 Flash only
- 5
- Like-for-like categories
- 6 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
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
Step 3.7 Flash
Step 3.7 Flash leads, 19.5 against 16.7.
Same board, both measured
Reasoning
GPQA Diamond
No call
Only o4 Mini has a GPQA Diamond result, 74.5%.
Directional only
Math
AIME
No call
Only o4 Mini has a AIME result, 83.7%.
Directional only
Knowledge
MMLU Pro
No call
Only o4 Mini has a MMLU Pro result, 80.6%.
Directional only
Web design
Design Arena: websites
Step 3.7 Flash
Step 3.7 Flash leads, 1196 against 985.
Same board, both measured
Long documents
Listed context window
Step 3.7 Flash
Step 3.7 Flash takes 262K tokens against 200K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Step 3.7 Flash
o4 Mini $0.0033, Step 3.7 Flash $0.0008 per turn.
List prices
At a glance
- Coding index
- o4 Mini not benchmarked, Step 3.7 Flash not benchmarked.
- Widest category
- Agentic work, 43.0 points apart on average over 1 shared benchmark.
- Widest single gap
- τ²-Bench Telecom: o4 Mini 55.6% against Step 3.7 Flash 98.5%, 43.0 points apart on a 0 to 100 scale.
- Output price
- o4 Mini $4.40 against Step 3.7 Flash $1.15 per million output tokens, 3.8x apart.
- Context window
- o4 Mini takes 200K tokens, Step 3.7 Flash 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 | o4 Mini | Bars | Step 3.7 Flash | Ties |
|---|---|---|---|---|
| Coding | 0 | 1 | ||
| Intelligence | 0 | 1 | ||
| Agentic work | 0 | 1 | ||
| Reasoning | 0 | 2 | ||
| Web design | 0 | 6 | ||
| General | 1 | 1 | ||
| All shared boards | 1 | 12 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index01
| Benchmark | o4 Mini | Step 3.7 Flash | 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 |
| AA Coding Index | Not measured | 39.6 | Only Step 3.7 Flash measured |
| LiveCodeBench (Vals AI) | 82.2% | Not measured | Only o4 Mini measured |
| LiveCodeBench (Artificial Analysis run) | 85.9% | Not measured | Only o4 Mini measured |
| SciCode | Not measured | 43.9% | Only Step 3.7 Flash measured |
| Terminal-Bench Hard (Artificial Analysis run) | 15.2% | 35.6% | Step 3.7 Flash |
| GSO-Bench | 3.6% | Not measured | Only o4 Mini measured |
| WeirdML | 52.6% | Not measured | Only o4 Mini measured |
| ALE-Bench | Not measured | 694.1 | Only Step 3.7 Flash measured |
| LMArena Text: coding | 1368 | Not measured | Only o4 Mini measured |
| LMArena Text: software and IT | 1367 | Not measured | Only o4 Mini measured |
| Neither model has a result on the other 25 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | o4 Mini | Step 3.7 Flash | Leader |
|---|---|---|---|
| AA Intelligence Index | 16.7 | 19.5 | Step 3.7 Flash |
| Epoch Capabilities Index | 145.5 | Not measured | Only o4 Mini measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index01
| Benchmark | o4 Mini | Step 3.7 Flash | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 55.6% | 98.5% | Step 3.7 Flash |
| METR time horizons | 63.9% | Not measured | Only o4 Mini measured |
| Neither model has a result on the other 37 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | o4 Mini | Step 3.7 Flash | Leader |
|---|---|---|---|
| GPQA Diamond | 74.5% | Not measured | Only o4 Mini measured |
| MMMU Pro (Vals AI) | 79.7% | Not measured | Only o4 Mini measured |
| GPQA Diamond (Artificial Analysis run) | 78.4% | 80.9% | Step 3.7 Flash |
| Humanity's Last Exam (Artificial Analysis run) | 16.5% | 21.4% | Step 3.7 Flash |
| GPQA Diamond (Epoch AI run) | 79.6% | Not measured | Only o4 Mini measured |
| ARC-AGI | 58.7% | Not measured | Only o4 Mini measured |
| ARC-AGI-2 | 6.1% | Not measured | Only o4 Mini measured |
| 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 | Not measured | Only o4 Mini measured |
| LMArena Text: expert | 1343 | Not measured | Only o4 Mini measured |
| LMArena Text: hard prompts in English | 1366 | Not measured | Only o4 Mini measured |
| LMArena Vision: diagrams | 1197 | Not measured | Only o4 Mini measured |
| LMArena Vision: homework | 1234 | Not measured | Only o4 Mini measured |
| Neither model has a result on the other 5 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index11 boards, none shared
| Benchmark | o4 Mini | Step 3.7 Flash | 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 |
| MATH level 5 | 97.8% | Not measured | Only o4 Mini measured |
| LMArena Text: math | 1389 | Not measured | Only o4 Mini measured |
| LMArena Text: mathematical work | 1379 | Not measured | Only o4 Mini measured |
| Neither model has a result on the other 2 boards here | |||
KnowledgeDirectional onlyPublic boards, shown as published, kept out of the index14 boards, none shared
| Benchmark | o4 Mini | Step 3.7 Flash | Leader |
|---|---|---|---|
| MMLU Pro | 80.6% | Not measured | Only o4 Mini measured |
| LegalBench (Vals AI) | 79.2% | Not measured | Only o4 Mini measured |
| CorpFin (Vals AI) | 59.0% | Not measured | Only o4 Mini measured |
| TaxEval (Vals AI) | 74.8% | Not measured | Only o4 Mini measured |
| MortgageTax (Vals AI) | 64.8% | Not measured | Only o4 Mini measured |
| MedQA (Vals AI) | 96.0% | Not measured | Only o4 Mini measured |
| MedCode (Vals AI) | 33.8% | Not measured | Only o4 Mini measured |
| 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 | Not measured | Only o4 Mini measured |
| LMArena Text: medicine and healthcare | 1361 | Not measured | Only o4 Mini measured |
| LMArena Text: legal and government | 1350 | Not measured | Only o4 Mini measured |
| LMArena Text: business and finance | 1325 | Not measured | Only o4 Mini measured |
| Neither model has a result on the other 3 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index06
| Benchmark | o4 Mini | Step 3.7 Flash | Leader |
|---|---|---|---|
| Design Arena: websites | 985 | 1196 | Step 3.7 Flash |
| Design Arena: UI components | 983 | 1179 | Step 3.7 Flash |
| Design Arena: game dev | 1003 | 1158 | Step 3.7 Flash |
| Design Arena: data viz | 990 | 1172 | Step 3.7 Flash |
| Design Arena: 3D | 854 | 1129 | Step 3.7 Flash |
| Design Arena: all code categories | 976 | 1182 | Step 3.7 Flash |
| Design Arena: SVG | Not measured | 1085 | Only Step 3.7 Flash measured |
| Design Arena: ASCII art | Not measured | 1152 | Only Step 3.7 Flash measured |
| Neither model has a result on the other 1 board here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index11
| Benchmark | o4 Mini | Step 3.7 Flash | Leader |
|---|---|---|---|
| SAGE (Vals AI) | 41.1% | Not measured | Only o4 Mini measured |
| MedScribe (Vals AI) | 69.1% | Not measured | Only o4 Mini measured |
| IFBench | 68.7% | 67.3% | o4 Mini |
| Long context reasoning (AA-LCR) | 61.0% | 73.7% | Step 3.7 Flash |
| DTBench | 77.6% | Not measured | Only o4 Mini measured |
| LMCA | 26.5% | Not measured | Only o4 Mini measured |
| Fiction.LiveBench | 77.8% | Not measured | Only o4 Mini measured |
| Creative writing (Lech Mazur) | 75.0% | Not measured | Only o4 Mini measured |
| LMArena Text | 1353 | Not measured | Only o4 Mini measured |
| LMArena Text: instruction following | 1320 | Not measured | Only o4 Mini measured |
| LMArena Text: creative writing | 1294 | Not measured | Only o4 Mini measured |
| LMArena Vision | 1194 | Not measured | Only o4 Mini measured |
| LMArena Text: English | 1367 | Not measured | Only o4 Mini measured |
| LMArena Text: non-English | 1337 | Not measured | Only o4 Mini measured |
| LMArena Text: Chinese | 1354 | Not measured | Only o4 Mini measured |
| LMArena Text: German | 1336 | Not measured | Only o4 Mini measured |
| LMArena Text: Japanese | 1308 | Not measured | Only o4 Mini measured |
| LMArena Text: Korean | 1312 | Not measured | Only o4 Mini measured |
| LMArena Text: French | 1364 | Not measured | Only o4 Mini measured |
| LMArena Text: Spanish | 1346 | Not measured | Only o4 Mini measured |
| LMArena Text: Russian | 1334 | Not measured | Only o4 Mini measured |
| LMArena Text: Polish | 1358 | Not measured | Only o4 Mini measured |
| LMArena Text: multi-turn | 1350 | Not measured | Only o4 Mini measured |
| LMArena Text: longer queries | 1315 | Not measured | Only o4 Mini measured |
| LMArena Text: writing and language | 1316 | Not measured | Only o4 Mini measured |
| LMArena Text: entertainment and media | 1308 | Not measured | Only o4 Mini measured |
| LMArena Vision: English | 1200 | Not measured | Only o4 Mini measured |
| LMArena Vision: Chinese | 1218 | Not measured | Only o4 Mini measured |
| LMArena Vision: reading text | 1197 | Not measured | Only o4 Mini measured |
| LMArena Vision: humor | 1199 | Not measured | Only o4 Mini measured |
| 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 | Not measured | Only o4 Mini measured |
| Neither model has a result on the other 13 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | o4 Mini | Step 3.7 Flash | 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
Step 3.7 Flash costs less
Repository review
50K input + 3K output tokens
Step 3.7 Flash costs less
Agent loop
200K input + 10K output tokens
Step 3.7 Flash 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 | StepFun Step 3.7 Flash |
|---|---|---|
| Maker | OpenAI | StepFun |
| Weights | No public weights listed | Open weights |
| Released | April 16, 2025 | May 28, 2026 |
| Context window | 200K | 262K |
| Input $/Mtok | $1.10 | $0.20 |
| Output $/Mtok | $4.40 | $1.15 |
| 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 o4 Mini better than Step 3.7 Flash for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. o4 Mini has the edge where what matters is Terminal-Bench Hard (Artificial Analysis run) or AA Intelligence Index, where it leads.
Which is cheaper, o4 Mini or Step 3.7 Flash?
o4 Mini costs $1.10 per million input tokens and $4.40 per million output tokens; Step 3.7 Flash costs $0.20 and $1.15. Cost per solved task is known for neither of them.
Which has the bigger context window, o4 Mini or Step 3.7 Flash?
Step 3.7 Flash, with 262K tokens against 200K.
How many benchmarks is the o4 Mini vs Step 3.7 Flash comparison based on?
13 benchmarks have a result for both models as of October 21, 2026. o4 Mini's coding index rests on 0 independent benchmarks and Step 3.7 Flash's on 0. Where only one model has a result, the page says so instead of naming a winner.