GPT-5.6 Sol vs Step 5 Preview
Updated October 21, 2026. 60 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
vs boards won of 60 shared
Decision reading
Pick GPT-5.6 Sol for coding: it scores 76.7 on the Vibe Coding Index against 71.4 for Step 5 Preview.
Pick Step 5 Preview when output tokens cost 7.4x less, or when what matters is SciCode or LiveBench research, where it leads.
Model AGPT-5.6 SolOpenAI76.7/1002 independent benchmarks$1.20 per solved taskModel BStep 5 PreviewStepFun71.4/1002 independent benchmarksNo cost per solved task
- Shared results
- 60
- GPT-5.6 Sol only
- 76
- Step 5 Preview only
- 1
- Like-for-like categories
- 8 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, four public coding benchmarks
GPT-5.6 Sol
GPT-5.6 Sol scores 76.7 against 71.4.
2 benchmarks behind each score
Intelligence
AA Intelligence Index
GPT-5.6 Sol
GPT-5.6 Sol leads, 47.0 against 43.7.
Same board, both measured
Agentic work
AA Agentic Index
No call
Only GPT-5.6 Sol has a AA Agentic Index result, 50.2.
Directional only
Reasoning
GPQA Diamond
No call
Only GPT-5.6 Sol has a GPQA Diamond result, 95.2%.
Directional only
Knowledge
MMLU Pro
No call
Only GPT-5.6 Sol has a MMLU Pro result, 89.1%.
Directional only
Web design
Design Arena: websites
No call
Only GPT-5.6 Sol has a Design Arena: websites result, 1321.
Directional only
Long documents
Listed context window
GPT-5.6 Sol
GPT-5.6 Sol takes 1.05M tokens against 1.05M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Step 5 Preview
GPT-5.6 Sol $0.014, Step 5 Preview $0.0024 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only GPT-5.6 Sol has the number, $1.20.
Directional only
At a glance
- Coding index
- GPT-5.6 Sol 76.7 from 2 benchmarks, Step 5 Preview 71.4 from 2 benchmarks.
- Widest category
- Agentic work, 17.8 points apart on average over 11 shared benchmarks.
- Widest single gap
- LiveBench browser: GPT-5.6 Sol 88.0% against Step 5 Preview 24.0%, 64.0 points apart on a 0 to 100 scale.
- Output price
- GPT-5.6 Sol $20.00 against Step 5 Preview $2.70 per million output tokens, 7.4x apart.
- Context window
- GPT-5.6 Sol takes 1.05M tokens, Step 5 Preview 1.05M.
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-5.6 Sol | Bars | Step 5 Preview | Ties |
|---|---|---|---|---|
| Coding | 11 | 1 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 10 | 1 | ||
| Reasoning | 7 | 1 | ||
| Math | 3 | 0 | ||
| Knowledge | 3 | 1 | ||
| Web design | 1 | 0 | ||
| General | 18 | 2 | ||
| All shared boards | 54 | 6 |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index111
| Benchmark | GPT-5.6 Sol | Step 5 Preview | Leader |
|---|---|---|---|
| LiveBench coding | 89.7% | 85.6% | GPT-5.6 Sol |
| LMArena WebDev | 1616 | 1567 | GPT-5.6 Sol |
| AA Coding Index | 77.4 | Not measured | Only GPT-5.6 Sol measured |
| LiveCodeBench (Vals AI) | 82.6% | Not measured | Only GPT-5.6 Sol measured |
| SWE-bench Verified (Vals AI) | 96.2% | Not measured | Only GPT-5.6 Sol measured |
| Terminal-Bench 4 (Vals AI) | 37.9% | 31.8% | GPT-5.6 Sol |
| Vibe Code Bench (Vals AI) | 80.5% | 70.4% | GPT-5.6 Sol |
| Code migration (Vals AI) | 52.9% | 39.0% | GPT-5.6 Sol |
| ProgramBench (Vals AI) | 1.5% | Not measured | Only GPT-5.6 Sol measured |
| IOI (Vals AI) | 91.2% | 38.1% | GPT-5.6 Sol |
| Terminal-Bench 2.1 (Vals AI) | 85.8% | Not measured | Only GPT-5.6 Sol measured |
| SWE-bench Multimodal (Vals AI) | Not measured | 40.8% | Only Step 5 Preview measured |
| Vibe Code Bench 1-100 (Vals AI) | 20.0% | Not measured | Only GPT-5.6 Sol measured |
| SkillsBench (Vals AI) | 54.1% | Not measured | Only GPT-5.6 Sol measured |
| CyberBench (Vals AI) | 76.3% | Not measured | Only GPT-5.6 Sol measured |
| SciCode | 57.8% | 58.9% | Step 5 Preview |
| Terminal-Bench Hard (Artificial Analysis run) | 65.9% | Not measured | Only GPT-5.6 Sol measured |
| FrontierCode | 47.5% | Not measured | Only GPT-5.6 Sol measured |
| FrontierSWE | 32.2% | Not measured | Only GPT-5.6 Sol measured |
| GSO-Bench | 76.5% | Not measured | Only GPT-5.6 Sol measured |
| WeirdML | 88.8% | Not measured | Only GPT-5.6 Sol measured |
| MirrorCode | 20.0% | Not measured | Only GPT-5.6 Sol measured |
| ALE-Bench | 2176.9 | Not measured | Only GPT-5.6 Sol measured |
| LMArena Text: coding | 1496 | 1485 | GPT-5.6 Sol |
| LMArena Text: software and IT | 1489 | 1473 | GPT-5.6 Sol |
| LMArena WebDev: React | 1620 | 1583 | GPT-5.6 Sol |
| LMArena WebDev: HTML | 1618 | 1556 | GPT-5.6 Sol |
| LMArena WebDev: from an image | 1609 | 1550 | GPT-5.6 Sol |
| Neither model has a result on the other 10 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | GPT-5.6 Sol | Step 5 Preview | Leader |
|---|---|---|---|
| AA Intelligence Index | 47.0 | 43.7 | GPT-5.6 Sol |
| Epoch Capabilities Index | 162.8 | Not measured | Only GPT-5.6 Sol measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index101
| Benchmark | GPT-5.6 Sol | Step 5 Preview | Leader |
|---|---|---|---|
| AA Agentic Index | 50.2 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: agents building web apps | 1245 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: agents building full-stack apps | 1261 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: agents building mobile apps | 1249 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: agents building Android apps | 1246 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: agents building Godot games | 1260 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: agents building games | 1267 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: HTML slides | 1286 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: PowerPoint slides | 1240 | Not measured | Only GPT-5.6 Sol measured |
| LiveBench research | 54.5% | 72.8% | Step 5 Preview |
| LiveBench browser | 88.0% | 24.0% | GPT-5.6 Sol |
| LiveBench long agentic | 48.5% | 27.5% | GPT-5.6 Sol |
| SREBench (Vals AI) | 30.5% | 0.4% | GPT-5.6 Sol |
| Finance Agent (Vals AI) | 53.8% | 50.7% | GPT-5.6 Sol |
| Tax Agent Bench (Vals AI) | 31.1% | 27.3% | GPT-5.6 Sol |
| Legal Research Bench (Vals AI) | 48.1% | 40.4% | GPT-5.6 Sol |
| Excel Modeling Benchmark (Vals AI) | 72.3% | 57.1% | GPT-5.6 Sol |
| BioMysteryBench (Vals AI) | 71.1% | 67.4% | GPT-5.6 Sol |
| Terminal-Bench-Science (Vals AI) | 20.0% | 2.9% | GPT-5.6 Sol |
| Vals Science Index | 41.5% | 29.6% | GPT-5.6 Sol |
| Time Horizon Index (Vals AI) | 23.8% | Not measured | Only GPT-5.6 Sol measured |
| CUA-bench (Vals AI) | 8.3% | Not measured | Only GPT-5.6 Sol measured |
| τ²-Bench Telecom | 85.1% | Not measured | Only GPT-5.6 Sol measured |
| OSWorld 2.0 | 27.3% | Not measured | Only GPT-5.6 Sol measured |
| BALROG | 60.0% | Not measured | Only GPT-5.6 Sol measured |
| PostTrainBench | 36.2% | Not measured | Only GPT-5.6 Sol measured |
| Vending-Bench 2 | $9,619 | Not measured | Only GPT-5.6 Sol measured |
| LMArena Search | 1261 | Not measured | Only GPT-5.6 Sol measured |
| Neither model has a result on the other 11 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index71
| Benchmark | GPT-5.6 Sol | Step 5 Preview | Leader |
|---|---|---|---|
| GPQA Diamond | 95.2% | Not measured | Only GPT-5.6 Sol measured |
| LiveBench reasoning | 91.7% | 83.8% | GPT-5.6 Sol |
| MysteryMechanism (Vals AI) | 33.3% | 18.5% | GPT-5.6 Sol |
| MMMU Pro (Vals AI) | 88.8% | Not measured | Only GPT-5.6 Sol measured |
| GPQA Diamond (Artificial Analysis run) | 94.1% | Not measured | Only GPT-5.6 Sol measured |
| Humanity's Last Exam (Artificial Analysis run) | 49.5% | 46.5% | GPT-5.6 Sol |
| GPQA Diamond (Epoch AI run) | 93.5% | Not measured | Only GPT-5.6 Sol measured |
| ARC-AGI | 97.5% | Not measured | Only GPT-5.6 Sol measured |
| ARC-AGI-2 | 92.5% | Not measured | Only GPT-5.6 Sol measured |
| SimpleBench | 64.8% | Not measured | Only GPT-5.6 Sol measured |
| Chess puzzles | 59.0% | Not measured | Only GPT-5.6 Sol measured |
| Mystery game puzzles | 58.0% | Not measured | Only GPT-5.6 Sol measured |
| EBR-bench | 44.8% | Not measured | Only GPT-5.6 Sol measured |
| LMArena Text: hard prompts | 1483 | 1461 | GPT-5.6 Sol |
| LMArena Text: expert | 1517 | 1485 | GPT-5.6 Sol |
| LMArena Text: hard prompts in English | 1477 | 1454 | GPT-5.6 Sol |
| LMArena Vision: diagrams | 1305 | 1308 | Step 5 Preview |
| LMArena Vision: homework | 1305 | 1265 | GPT-5.6 Sol |
| Neither model has a result on the other 3 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | GPT-5.6 Sol | Step 5 Preview | Leader |
|---|---|---|---|
| OTIS Mock AIME | 100.0% | Not measured | Only GPT-5.6 Sol measured |
| FrontierMath, tiers 1 to 3 | 89.1% | Not measured | Only GPT-5.6 Sol measured |
| FrontierMath, tier 4 | 82.9% | Not measured | Only GPT-5.6 Sol measured |
| ProofBench | 83.0% | 42.0% | GPT-5.6 Sol |
| LMArena Text: math | 1473 | 1458 | GPT-5.6 Sol |
| LMArena Text: mathematical work | 1496 | 1460 | GPT-5.6 Sol |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index31
| Benchmark | GPT-5.6 Sol | Step 5 Preview | Leader |
|---|---|---|---|
| MMLU Pro | 89.1% | Not measured | Only GPT-5.6 Sol measured |
| LegalBench (Vals AI) | 87.0% | Not measured | Only GPT-5.6 Sol measured |
| CorpFin (Vals AI) | 64.4% | Not measured | Only GPT-5.6 Sol measured |
| TaxEval (Vals AI) | 74.8% | Not measured | Only GPT-5.6 Sol measured |
| MortgageTax (Vals AI) | 67.3% | Not measured | Only GPT-5.6 Sol measured |
| MedCode (Vals AI) | 44.0% | Not measured | Only GPT-5.6 Sol measured |
| SimpleQA Verified | 69.7% | Not measured | Only GPT-5.6 Sol measured |
| LMArena Text: science | 1455 | 1454 | GPT-5.6 Sol |
| LMArena Text: medicine and healthcare | 1431 | 1439 | Step 5 Preview |
| LMArena Text: legal and government | 1476 | 1437 | GPT-5.6 Sol |
| LMArena Text: business and finance | 1457 | 1435 | GPT-5.6 Sol |
| Neither model has a result on the other 6 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | GPT-5.6 Sol | Step 5 Preview | Leader |
|---|---|---|---|
| Design Arena: websites | 1321 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: UI components | 1338 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: game dev | 1368 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: data viz | 1323 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: 3D | 1385 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: all code categories | 1341 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: SVG | 1316 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: ASCII art | 1302 | Not measured | Only GPT-5.6 Sol measured |
| LiveBench design | 81.6% | 76.9% | GPT-5.6 Sol |
GeneralLike-for-likePublic boards, shown as published, kept out of the index182
| Benchmark | GPT-5.6 Sol | Step 5 Preview | Leader |
|---|---|---|---|
| LiveBench data analysis | 79.8% | 63.5% | GPT-5.6 Sol |
| Public Benefits Bench (Vals AI) | 66.5% | 58.1% | GPT-5.6 Sol |
| SAGE (Vals AI) | 52.6% | Not measured | Only GPT-5.6 Sol measured |
| MedScribe (Vals AI) | 85.2% | Not measured | Only GPT-5.6 Sol measured |
| Vals Multimodal Index | 72.6% | Not measured | Only GPT-5.6 Sol measured |
| IFBench | 72.7% | Not measured | Only GPT-5.6 Sol measured |
| Long context reasoning (AA-LCR) | 84.0% | 88.3% | Step 5 Preview |
| DTBench | 95.7% | Not measured | Only GPT-5.6 Sol measured |
| LMCA | 58.5% | Not measured | Only GPT-5.6 Sol measured |
| Surface Evolver Bench | 93.1% | Not measured | Only GPT-5.6 Sol measured |
| Furniture assembly | 56.7% | Not measured | Only GPT-5.6 Sol measured |
| LMArena Text | 1457 | 1440 | GPT-5.6 Sol |
| LMArena Text: instruction following | 1483 | 1438 | GPT-5.6 Sol |
| LMArena Text: creative writing | 1449 | 1409 | GPT-5.6 Sol |
| LMArena Vision | 1281 | 1265 | GPT-5.6 Sol |
| LMArena Text: English | 1457 | 1447 | GPT-5.6 Sol |
| LMArena Text: non-English | 1451 | 1428 | GPT-5.6 Sol |
| LMArena Text: Chinese | 1525 | 1500 | GPT-5.6 Sol |
| LMArena Text: German | 1476 | Not measured | Only GPT-5.6 Sol measured |
| LMArena Text: Japanese | 1472 | 1383 | GPT-5.6 Sol |
| LMArena Text: Korean | 1442 | Not measured | Only GPT-5.6 Sol measured |
| LMArena Text: French | 1472 | 1438 | GPT-5.6 Sol |
| LMArena Text: Spanish | 1443 | 1441 | GPT-5.6 Sol |
| LMArena Text: Russian | 1471 | 1433 | GPT-5.6 Sol |
| LMArena Text: Polish | 1459 | Not measured | Only GPT-5.6 Sol measured |
| LMArena Text: multi-turn | 1461 | 1437 | GPT-5.6 Sol |
| LMArena Text: longer queries | 1481 | 1449 | GPT-5.6 Sol |
| LMArena Text: writing and language | 1462 | 1425 | GPT-5.6 Sol |
| LMArena Text: entertainment and media | 1450 | 1412 | GPT-5.6 Sol |
| LMArena Vision: English | 1280 | 1290 | Step 5 Preview |
| LMArena Vision: Chinese | 1313 | Not measured | Only GPT-5.6 Sol measured |
| LMArena Vision: reading text | 1290 | 1282 | GPT-5.6 Sol |
| LMArena Vision: humor | 1246 | Not measured | Only GPT-5.6 Sol measured |
| LMArena Vision: creative writing | 1283 | Not measured | Only GPT-5.6 Sol measured |
| LMArena Document | 1483 | Not measured | Only GPT-5.6 Sol measured |
| 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 | GPT-5.6 Sol | Step 5 Preview | 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-5.6 Sol$0.014Fits in one request
Step 5 Preview$0.0024Fits in one request
Step 5 Preview costs less
Repository review
50K input + 3K output tokens
GPT-5.6 Sol$0.260Fits in one request
Step 5 Preview$0.058Fits in one request
Step 5 Preview costs less
Agent loop
200K input + 10K output tokens
GPT-5.6 Sol$1.00Fits in one request
Step 5 Preview$0.227Fits in one request
Step 5 Preview 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-5.6 Sol | StepFun Step 5 Preview |
|---|---|---|
| Maker | OpenAI | StepFun |
| Weights | No public weights listed | Open weights |
| Released | July 9, 2026 | October 8, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $4.00 | $1.00 |
| Output $/Mtok | $20.00 | $2.70 |
| Coding index | 76.7 | 71.4 |
| Independent measurements behind it | 2 | 2 |
| Cost per solved task | $1.20 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is GPT-5.6 Sol better than Step 5 Preview for coding?
Pick GPT-5.6 Sol for coding: it scores 76.7 on the Vibe Coding Index against 71.4 for Step 5 Preview. Pick Step 5 Preview when output tokens cost 7.4x less, or when what matters is SciCode or LiveBench research, where it leads.
Which is cheaper, GPT-5.6 Sol or Step 5 Preview?
GPT-5.6 Sol costs $4.00 per million input tokens and $20.00 per million output tokens; Step 5 Preview costs $1.00 and $2.70. Cost per solved task is known for GPT-5.6 Sol only.
Which has the bigger context window, GPT-5.6 Sol or Step 5 Preview?
GPT-5.6 Sol, with 1.05M tokens against 1.05M.
How many benchmarks is the GPT-5.6 Sol vs Step 5 Preview comparison based on?
60 benchmarks have a result for both models as of October 21, 2026. GPT-5.6 Sol's coding index rests on 2 independent benchmarks and Step 5 Preview's on 2. Where only one model has a result, the page says so instead of naming a winner.