GPT-5.6 Sol vs Grok 4.7
GPT-5.6 Sol leads Grok 4.7 on coding, 73.5 to 66.7 on the Vibe Coding Index, and Grok 4.7 costs 3.3x less per output token ($6.00 vs $20.00 per million). Updated October 21, 2026; 94 benchmarks have a result for both.
GPT-5.6 Sol or Grok 4.7: which to pick
Pick GPT-5.6 Sol for coding: it scores 73.5 on the Vibe Coding Index against 66.7 for Grok 4.7.
Pick Grok 4.7 for output tokens 3.3x cheaper ($6.00 vs $20.00 per million) and work measured by LMArena WebDev or Vibe Code Bench (Vals AI), where it leads.
- Shared results
- 94
- GPT-5.6 Sol only
- 45
- Grok 4.7 only
- 5
- Like-for-like categories
- 8 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
GPT-5.6 Sol
GPT-5.6 Sol scores 73.5 against 66.7.
6 and 5 benchmarks behind the scores
Intelligence
AA Intelligence Index
GPT-5.6 Sol
GPT-5.6 Sol leads, 47.0 against 46.4.
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
GPT-5.6 Sol
GPT-5.6 Sol leads, 1321 against 1254.
Same board, both measured
General
Vals Index
No call
Only Grok 4.7 has a Vals Index result, 58.6%.
Directional only
Long documents
Listed context window
GPT-5.6 Sol
GPT-5.6 Sol takes 1.05M tokens against 500K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Grok 4.7
GPT-5.6 Sol $0.014, Grok 4.7 $0.0050 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 73.5 from 6 benchmarks, Grok 4.7 66.7 from 5 benchmarks.
- Widest category
- Math, 30.2 points apart on average over 6 shared benchmarks.
- Widest single gap
- FrontierMath, tier 4: GPT-5.6 Sol 82.9% against Grok 4.7 17.1%, 65.9 points apart on a 0 to 100 scale.
- Output price
- GPT-5.6 Sol $20.00 against Grok 4.7 $6.00 per million output tokens, 3.3x apart.
- Context window
- GPT-5.6 Sol takes 1.05M tokens, Grok 4.7 500K.
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 | Grok 4.7 | Ties |
|---|---|---|---|---|
| Coding | 11 | 5 | 1 tie | |
| Intelligence | 2 | 0 | ||
| Agentic work | 8 | 4 | ||
| Reasoning | 12 | 1 | ||
| Math | 6 | 0 | ||
| Knowledge | 6 | 1 | ||
| Web design | 7 | 0 | ||
| General | 28 | 2 | ||
| All shared boards | 80 | 13 | 1 tie |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index115
| Benchmark | GPT-5.6 Sol | Grok 4.7 | Leader |
|---|---|---|---|
| LiveBench coding | 89.7% | 84.5% | GPT-5.6 Sol |
| LMArena WebDev | 1616 | 1637 | Grok 4.7 |
| 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% | 28.8% | GPT-5.6 Sol |
| Vibe Code Bench (Vals AI) | 80.5% | 86.2% | Grok 4.7 |
| Code migration (Vals AI) | 52.9% | 44.8% | GPT-5.6 Sol |
| ProgramBench (Vals AI) | 1.5% | 0.5% | GPT-5.6 Sol |
| IOI (Vals AI) | 91.2% | 57.7% | GPT-5.6 Sol |
| Terminal-Bench 2.1 (Vals AI) | 85.8% | 73.4% | GPT-5.6 Sol |
| SWE-bench Multimodal (Vals AI) | Not measured | 42.9% | Only Grok 4.7 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% | 69.5% | GPT-5.6 Sol |
| SciCode | 57.8% | 57.8% | Tie |
| Terminal-Bench Hard (Artificial Analysis run) | 65.9% | Not measured | Only GPT-5.6 Sol measured |
| FrontierCode | 47.5% | 47.6% | Grok 4.7 |
| FrontierSWE | 32.2% | 29.5% | GPT-5.6 Sol |
| 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 | 1432 | GPT-5.6 Sol |
| LMArena Text: software and IT | 1489 | 1421 | GPT-5.6 Sol |
| LMArena WebDev: React | 1620 | 1663 | Grok 4.7 |
| LMArena WebDev: HTML | 1618 | 1627 | Grok 4.7 |
| LMArena WebDev: from an image | 1609 | 1586 | GPT-5.6 Sol |
| SciCode | 10.8% | Not measured | Only GPT-5.6 Sol measured |
| Neither model has a result on the other 10 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | GPT-5.6 Sol | Grok 4.7 | Leader |
|---|---|---|---|
| AA Intelligence Index | 47.0 | 46.4 | GPT-5.6 Sol |
| Epoch Capabilities Index | 162.8 | 154.4 | GPT-5.6 Sol |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index84
| Benchmark | GPT-5.6 Sol | Grok 4.7 | 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 | 1248 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: agents building Android apps | 1239 | 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 | 1241 | Not measured | Only GPT-5.6 Sol measured |
| LiveBench research | 54.5% | 86.4% | Grok 4.7 |
| LiveBench browser | 88.0% | 60.0% | GPT-5.6 Sol |
| LiveBench long agentic | 48.5% | 50.1% | Grok 4.7 |
| SREBench (Vals AI) | 30.5% | 3.8% | GPT-5.6 Sol |
| Finance Agent (Vals AI) | 53.8% | 52.3% | GPT-5.6 Sol |
| Tax Agent Bench (Vals AI) | 31.1% | 34.2% | Grok 4.7 |
| Legal Research Bench (Vals AI) | 48.1% | 47.1% | GPT-5.6 Sol |
| Harvey's Legal Agent Benchmark (Vals AI) | Not measured | 18.3% | Only Grok 4.7 measured |
| Excel Modeling Benchmark (Vals AI) | 72.3% | 67.0% | GPT-5.6 Sol |
| BioMysteryBench (Vals AI) | 71.1% | 69.3% | GPT-5.6 Sol |
| Terminal-Bench-Science (Vals AI) | 20.0% | 10.0% | GPT-5.6 Sol |
| Vals Science Index | 41.5% | 34.8% | 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 |
| Robo-bench (Vals AI) | Not measured | 0.2% | Only Grok 4.7 measured |
| τ²-Bench Telecom | 85.1% | Not measured | Only GPT-5.6 Sol measured |
| APEX-Agents | Not measured | 54.6% | Only Grok 4.7 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 | $10,537 | Grok 4.7 |
| LMArena Search | 1261 | Not measured | Only GPT-5.6 Sol measured |
| Neither model has a result on the other 10 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index121
| Benchmark | GPT-5.6 Sol | Grok 4.7 | Leader |
|---|---|---|---|
| GPQA Diamond | 95.2% | Not measured | Only GPT-5.6 Sol measured |
| LiveBench reasoning | 91.7% | 82.7% | GPT-5.6 Sol |
| MysteryMechanism (Vals AI) | 33.3% | 25.2% | 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% | 43.1% | GPT-5.6 Sol |
| GPQA Diamond (Epoch AI run) | 93.5% | 92.7% | GPT-5.6 Sol |
| ARC-AGI | 97.5% | 90.2% | GPT-5.6 Sol |
| ARC-AGI-2 | 92.5% | 61.4% | GPT-5.6 Sol |
| SimpleBench | 64.8% | Not measured | Only GPT-5.6 Sol measured |
| Chess puzzles | 59.0% | 41.0% | GPT-5.6 Sol |
| Mystery game puzzles | 58.0% | 29.0% | GPT-5.6 Sol |
| EBR-bench | 44.8% | Not measured | Only GPT-5.6 Sol measured |
| LMArena Text: hard prompts | 1483 | 1415 | GPT-5.6 Sol |
| LMArena Text: expert | 1517 | 1439 | GPT-5.6 Sol |
| LMArena Text: hard prompts in English | 1477 | 1427 | GPT-5.6 Sol |
| LMArena Vision: diagrams | 1305 | 1254 | GPT-5.6 Sol |
| LMArena Vision: homework | 1305 | 1309 | Grok 4.7 |
| Kagi LLM Benchmark | 67.0% | Not measured | Only GPT-5.6 Sol measured |
| Neither model has a result on the other 3 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | GPT-5.6 Sol | Grok 4.7 | Leader |
|---|---|---|---|
| OTIS Mock AIME | 100.0% | 98.1% | GPT-5.6 Sol |
| FrontierMath, tiers 1 to 3 | 89.1% | 53.0% | GPT-5.6 Sol |
| FrontierMath, tier 4 | 82.9% | 17.1% | GPT-5.6 Sol |
| ProofBench | 83.0% | 26.0% | GPT-5.6 Sol |
| LMArena Text: math | 1473 | 1414 | GPT-5.6 Sol |
| LMArena Text: mathematical work | 1496 | 1435 | 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 index61
| Benchmark | GPT-5.6 Sol | Grok 4.7 | Leader |
|---|---|---|---|
| MMLU Pro | 89.1% | Not measured | Only GPT-5.6 Sol measured |
| LegalBench (Vals AI) | 87.0% | 84.4% | GPT-5.6 Sol |
| 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% | 49.5% | Grok 4.7 |
| SimpleQA Verified | 69.7% | 56.0% | GPT-5.6 Sol |
| LMArena Text: science | 1455 | 1419 | GPT-5.6 Sol |
| LMArena Text: medicine and healthcare | 1431 | 1395 | GPT-5.6 Sol |
| LMArena Text: legal and government | 1476 | 1423 | GPT-5.6 Sol |
| LMArena Text: business and finance | 1457 | 1389 | GPT-5.6 Sol |
| FACTS Benchmark Suite | 52.4% | Not measured | Only GPT-5.6 Sol measured |
| Neither model has a result on the other 6 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index70
| Benchmark | GPT-5.6 Sol | Grok 4.7 | Leader |
|---|---|---|---|
| Design Arena: websites | 1321 | 1254 | GPT-5.6 Sol |
| Design Arena: UI components | 1336 | 1275 | GPT-5.6 Sol |
| Design Arena: game dev | 1368 | 1313 | GPT-5.6 Sol |
| Design Arena: data viz | 1322 | 1235 | GPT-5.6 Sol |
| Design Arena: 3D | 1386 | 1270 | GPT-5.6 Sol |
| Design Arena: all code categories | 1341 | 1271 | GPT-5.6 Sol |
| Design Arena: SVG | 1316 | Not measured | Only GPT-5.6 Sol measured |
| Design Arena: ASCII art | 1304 | Not measured | Only GPT-5.6 Sol measured |
| LiveBench design | 81.6% | 78.0% | GPT-5.6 Sol |
GeneralLike-for-likePublic boards, shown as published, kept out of the index282
| Benchmark | GPT-5.6 Sol | Grok 4.7 | Leader |
|---|---|---|---|
| Vals Index | Not measured | 58.6% | Only Grok 4.7 measured |
| LiveBench data analysis | 79.8% | 76.9% | GPT-5.6 Sol |
| Public Benefits Bench (Vals AI) | 66.5% | 65.6% | GPT-5.6 Sol |
| SAGE (Vals AI) | 52.6% | 40.8% | GPT-5.6 Sol |
| MedScribe (Vals AI) | 85.2% | 89.4% | Grok 4.7 |
| 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% | 80.0% | GPT-5.6 Sol |
| DTBench | 95.7% | 96.0% | Grok 4.7 |
| LMCA | 58.5% | 49.4% | GPT-5.6 Sol |
| Surface Evolver Bench | 93.1% | Not measured | Only GPT-5.6 Sol measured |
| Furniture assembly | 56.7% | 20.8% | GPT-5.6 Sol |
| LMArena Text | 1457 | 1399 | GPT-5.6 Sol |
| LMArena Text: instruction following | 1483 | 1406 | GPT-5.6 Sol |
| LMArena Text: creative writing | 1449 | 1381 | GPT-5.6 Sol |
| LMArena Vision | 1281 | 1231 | GPT-5.6 Sol |
| LMArena Text: English | 1457 | 1411 | GPT-5.6 Sol |
| LMArena Text: non-English | 1451 | 1386 | GPT-5.6 Sol |
| LMArena Text: Chinese | 1525 | 1449 | GPT-5.6 Sol |
| LMArena Text: German | 1476 | 1366 | GPT-5.6 Sol |
| LMArena Text: Japanese | 1472 | 1359 | GPT-5.6 Sol |
| LMArena Text: Korean | 1442 | 1383 | GPT-5.6 Sol |
| LMArena Text: French | 1472 | 1431 | GPT-5.6 Sol |
| LMArena Text: Spanish | 1443 | 1406 | GPT-5.6 Sol |
| LMArena Text: Russian | 1471 | 1399 | GPT-5.6 Sol |
| LMArena Text: Polish | 1459 | 1420 | GPT-5.6 Sol |
| LMArena Text: multi-turn | 1461 | 1394 | GPT-5.6 Sol |
| LMArena Text: longer queries | 1481 | 1416 | GPT-5.6 Sol |
| LMArena Text: writing and language | 1462 | 1394 | GPT-5.6 Sol |
| LMArena Text: entertainment and media | 1450 | 1376 | GPT-5.6 Sol |
| LMArena Vision: English | 1280 | 1229 | GPT-5.6 Sol |
| LMArena Vision: Chinese | 1313 | 1250 | GPT-5.6 Sol |
| LMArena Vision: reading text | 1290 | 1241 | GPT-5.6 Sol |
| LMArena Vision: humor | 1246 | Not measured | Only GPT-5.6 Sol measured |
| LMArena Vision: creative writing | 1283 | 1217 | GPT-5.6 Sol |
| LMArena Document | 1483 | Not measured | Only GPT-5.6 Sol measured |
| Neither model has a result on the other 10 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GPT-5.6 Sol | Grok 4.7 | 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
Grok 4.7 costs less
Repository review
50K input + 3K output tokens
Grok 4.7 costs less
Agent loop
200K input + 10K output tokens
Grok 4.7 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 | SpaceXAI Grok 4.7 |
|---|---|---|
| Maker | OpenAI | xAI |
| Weights | No public weights listed | No public weights listed |
| Released | July 9, 2026 | September 21, 2026 |
| Context window | 1.05M | 500K |
| Input $/Mtok | $4.00 | $2.00 |
| Output $/Mtok | $20.00 | $6.00 |
| Coding index | 73.5 | 66.7 |
| Independent measurements behind it | 6 | 5 |
| Cost per solved task | $1.20 | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is GPT-5.6 Sol better than Grok 4.7 for coding?
Pick GPT-5.6 Sol for coding: it scores 73.5 on the Vibe Coding Index against 66.7 for Grok 4.7. Pick Grok 4.7 for output tokens 3.3x cheaper ($6.00 vs $20.00 per million) and work measured by LMArena WebDev or Vibe Code Bench (Vals AI), where it leads.
Which is cheaper, GPT-5.6 Sol or Grok 4.7?
GPT-5.6 Sol costs $4.00 per million input tokens and $20.00 per million output tokens; Grok 4.7 costs $2.00 and $6.00. Cost per solved task is known for GPT-5.6 Sol only.
Which has the bigger context window, GPT-5.6 Sol or Grok 4.7?
GPT-5.6 Sol, with 1.05M tokens against 500K.
How many benchmarks is the GPT-5.6 Sol vs Grok 4.7 comparison based on?
94 benchmarks have a result for both models as of October 21, 2026. GPT-5.6 Sol's coding index rests on 6 independent benchmarks and Grok 4.7's on 5. Where only one model has a result, the page says so instead of naming a winner.