Grok 4.5 vs Grok 4.6
Grok 4.6 leads Grok 4.5 on coding, 65.7 to 58.6 on the Vibe Coding Index, at about the same output price ($6.00 vs $6.00 per million). Updated October 22, 2026; 114 benchmarks have a result for both.
Grok 4.5 or Grok 4.6: which to pick
Pick Grok 4.6 for coding: it scores 65.7 on the Vibe Coding Index against 58.6 for Grok 4.5.
Pick Grok 4.5 for a lower cost per solved task ($0.62 against $0.82) and work measured by LiveBench coding or SkillsBench (Vals AI), where it leads.
- Shared results
- 114
- Grok 4.5 only
- 9
- Grok 4.6 only
- 9
- 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
Grok 4.6
Grok 4.6 scores 65.7 against 58.6.
6 benchmarks behind each score
Intelligence
AA Intelligence Index
Grok 4.6
Grok 4.6 leads, 44.3 against 38.8.
Same board, both measured
Agentic work
AA Agentic Index
Grok 4.6
Grok 4.6 leads, 53.0 against 41.2.
Same board, both measured
Reasoning
GPQA Diamond
Grok 4.6
Grok 4.6 leads, 94.7% against 92.9%.
Same board, both measured
Knowledge
MMLU Pro
Grok 4.6
Grok 4.6 leads, 89.4% against 89.2%.
Same board, both measured
Web design
Design Arena: websites
Grok 4.6
Grok 4.6 leads, 1289 against 1281.
Same board, both measured
Long documents
Listed context window
No call
Both take 500K tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
No call
Grok 4.5 $0.0050, Grok 4.6 $0.0050 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
Grok 4.5
Grok 4.5 $0.62, Grok 4.6 $0.82.
Published attempt cost and pass rate
At a glance
- Coding index
- Grok 4.5 58.6 from 6 benchmarks, Grok 4.6 65.7 from 6 benchmarks.
- Widest category
- Agentic work, 8.0 points apart on average over 19 shared benchmarks.
- Widest single gap
- Vending-Bench 2: Grok 4.5 $3,887 against Grok 4.6 $9,047, 32.2 points apart on a 0 to 100 scale.
- Output price
- Both cost $6.00 per million output tokens.
- Context window
- Both take 500K tokens.
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 | Grok 4.5 | Bars | Grok 4.6 | Ties |
|---|---|---|---|---|
| Coding | 5 | 17 | ||
| Intelligence | 0 | 2 | ||
| Agentic work | 3 | 16 | ||
| Reasoning | 5 | 9 | ||
| Math | 2 | 4 | ||
| Knowledge | 6 | 5 | ||
| Web design | 0 | 9 | ||
| General | 25 | 6 | ||
| All shared boards | 46 | 68 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index517
| Benchmark | Grok 4.5 | Grok 4.6 | Leader |
|---|---|---|---|
| LiveBench coding | 94.3% | 89.1% | Grok 4.5 |
| LMArena WebDev | 1553 | 1617 | Grok 4.6 |
| AA Coding Index | 72.4 | 76.8 | Grok 4.6 |
| LiveCodeBench (Vals AI) | 87.3% | 88.2% | Grok 4.6 |
| SWE-bench Verified (Vals AI) | 86.6% | 95.6% | Grok 4.6 |
| Terminal-Bench 4 (Vals AI) | 8.6% | 17.2% | Grok 4.6 |
| Vibe Code Bench (Vals AI) | 69.0% | 76.2% | Grok 4.6 |
| Code migration (Vals AI) | 36.6% | 44.6% | Grok 4.6 |
| ProgramBench (Vals AI) | 0.0% | 1.0% | Grok 4.6 |
| IOI (Vals AI) | 40.6% | 47.6% | Grok 4.6 |
| Terminal-Bench 2.1 (Vals AI) | 67.8% | 78.3% | Grok 4.6 |
| Vibe Code Bench 1-100 (Vals AI) | Not measured | 14.8% | Only Grok 4.6 measured |
| SkillsBench (Vals AI) | 66.0% | 55.8% | Grok 4.5 |
| CyberBench (Vals AI) | 71.0% | 67.7% | Grok 4.5 |
| SciCode | 55.0% | 56.5% | Grok 4.6 |
| FrontierCode | 42.4% | 48.0% | Grok 4.6 |
| FrontierSWE | Not measured | 25.3% | Only Grok 4.6 measured |
| WeirdML | 46.4% | 67.3% | Grok 4.6 |
| ALE-Bench | 1309.3 | 1508.1 | Grok 4.6 |
| LMArena Text: coding | 1474 | 1466 | Grok 4.5 |
| LMArena Text: software and IT | 1471 | 1457 | Grok 4.5 |
| LMArena WebDev: React | 1556 | 1631 | Grok 4.6 |
| LMArena WebDev: HTML | 1567 | 1597 | Grok 4.6 |
| LMArena WebDev: from an image | 1562 | 1603 | Grok 4.6 |
| SciCode | 6.5% | Not measured | Only Grok 4.5 measured |
| Neither model has a result on the other 14 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Grok 4.5 | Grok 4.6 | Leader |
|---|---|---|---|
| AA Intelligence Index | 38.8 | 44.3 | Grok 4.6 |
| Epoch Capabilities Index | 154.1 | 156.6 | Grok 4.6 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index316
| Benchmark | Grok 4.5 | Grok 4.6 | Leader |
|---|---|---|---|
| AA Agentic Index | 41.2 | 53.0 | Grok 4.6 |
| Design Arena: agents building web apps | 1218 | 1248 | Grok 4.6 |
| Design Arena: agents building full-stack apps | 1249 | 1264 | Grok 4.6 |
| Design Arena: agents building mobile apps | 1220 | 1234 | Grok 4.6 |
| Design Arena: agents building Android apps | 1236 | 1291 | Grok 4.6 |
| Design Arena: agents building Godot games | 1229 | 1276 | Grok 4.6 |
| Design Arena: agents building games | 1236 | 1209 | Grok 4.5 |
| Design Arena: HTML slides | 1205 | 1250 | Grok 4.6 |
| Design Arena: PowerPoint slides | 1230 | 1237 | Grok 4.6 |
| LiveBench research | 85.9% | 84.5% | Grok 4.5 |
| LiveBench browser | 44.0% | 64.0% | Grok 4.6 |
| LiveBench long agentic | 27.9% | 40.7% | Grok 4.6 |
| SREBench (Vals AI) | 0.8% | Not measured | Only Grok 4.5 measured |
| Finance Agent (Vals AI) | 48.4% | 53.7% | Grok 4.6 |
| Tax Agent Bench (Vals AI) | 24.9% | 33.1% | Grok 4.6 |
| Legal Research Bench (Vals AI) | 38.0% | 48.1% | Grok 4.6 |
| Excel Modeling Benchmark (Vals AI) | 52.9% | 62.7% | Grok 4.6 |
| BioMysteryBench (Vals AI) | Not measured | 72.2% | Only Grok 4.6 measured |
| Terminal-Bench-Science (Vals AI) | Not measured | 11.4% | Only Grok 4.6 measured |
| Vals Science Index | Not measured | 38.1% | Only Grok 4.6 measured |
| Time Horizon Index (Vals AI) | 6.3% | 5.8% | Grok 4.5 |
| APEX-Agents | 56.2% | 65.3% | Grok 4.6 |
| PostTrainBench | 23.4% | Not measured | Only Grok 4.5 measured |
| Vending-Bench 2 | $3,887 | $9,047 | Grok 4.6 |
| LMArena Search | 1216 | Not measured | Only Grok 4.5 measured |
| Neither model has a result on the other 16 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index59
| Benchmark | Grok 4.5 | Grok 4.6 | Leader |
|---|---|---|---|
| GPQA Diamond | 92.9% | 94.7% | Grok 4.6 |
| LiveBench reasoning | 87.2% | 91.2% | Grok 4.6 |
| MysteryMechanism (Vals AI) | Not measured | 30.6% | Only Grok 4.6 measured |
| MMMU Pro (Vals AI) | 61.8% | Not measured | Only Grok 4.5 measured |
| GPQA Diamond (Artificial Analysis run) | 93.1% | 94.9% | Grok 4.6 |
| Humanity's Last Exam (Artificial Analysis run) | 42.7% | 44.1% | Grok 4.6 |
| GPQA Diamond (Epoch AI run) | 93.4% | 94.0% | Grok 4.6 |
| ARC-AGI | 87.2% | 87.5% | Grok 4.6 |
| ARC-AGI-2 | 52.6% | 67.1% | Grok 4.6 |
| SimpleBench | 70.0% | 75.9% | Grok 4.6 |
| Chess puzzles | 35.0% | 33.0% | Grok 4.5 |
| Mystery game puzzles | Not measured | 34.0% | Only Grok 4.6 measured |
| EBR-bench | Not measured | 30.5% | Only Grok 4.6 measured |
| LMArena Text: hard prompts | 1462 | 1448 | Grok 4.5 |
| LMArena Text: expert | 1465 | 1469 | Grok 4.6 |
| LMArena Text: hard prompts in English | 1464 | 1445 | Grok 4.5 |
| LMArena Vision: diagrams | 1304 | 1285 | Grok 4.5 |
| LMArena Vision: homework | 1295 | 1283 | Grok 4.5 |
| Kagi LLM Benchmark | 83.5% | Not measured | Only Grok 4.5 measured |
| Neither model has a result on the other 3 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index24
| Benchmark | Grok 4.5 | Grok 4.6 | Leader |
|---|---|---|---|
| OTIS Mock AIME | 97.8% | 99.2% | Grok 4.6 |
| FrontierMath, tiers 1 to 3 | 57.2% | 66.0% | Grok 4.6 |
| FrontierMath, tier 4 | 24.4% | 31.7% | Grok 4.6 |
| ProofBench | 31.0% | 51.0% | Grok 4.6 |
| LMArena Text: math | 1460 | 1417 | Grok 4.5 |
| LMArena Text: mathematical work | 1482 | 1464 | Grok 4.5 |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index65
| Benchmark | Grok 4.5 | Grok 4.6 | Leader |
|---|---|---|---|
| MMLU Pro | 89.2% | 89.4% | Grok 4.6 |
| LegalBench (Vals AI) | 86.0% | 86.3% | Grok 4.6 |
| CorpFin (Vals AI) | 67.4% | 66.2% | Grok 4.5 |
| TaxEval (Vals AI) | 71.7% | 71.1% | Grok 4.5 |
| MortgageTax (Vals AI) | 61.8% | 64.2% | Grok 4.6 |
| MedCode (Vals AI) | 43.3% | 44.7% | Grok 4.6 |
| SimpleQA Verified | 48.3% | 49.3% | Grok 4.6 |
| LMArena Text: science | 1464 | 1437 | Grok 4.5 |
| LMArena Text: medicine and healthcare | 1440 | 1420 | Grok 4.5 |
| LMArena Text: legal and government | 1463 | 1429 | Grok 4.5 |
| LMArena Text: business and finance | 1441 | 1416 | Grok 4.5 |
| FACTS Benchmark Suite | 50.2% | Not measured | Only Grok 4.5 measured |
| Neither model has a result on the other 6 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index09
| Benchmark | Grok 4.5 | Grok 4.6 | Leader |
|---|---|---|---|
| Design Arena: websites | 1281 | 1289 | Grok 4.6 |
| Design Arena: UI components | 1280 | 1294 | Grok 4.6 |
| Design Arena: game dev | 1264 | 1296 | Grok 4.6 |
| Design Arena: data viz | 1266 | 1282 | Grok 4.6 |
| Design Arena: 3D | 1257 | 1277 | Grok 4.6 |
| Design Arena: all code categories | 1275 | 1292 | Grok 4.6 |
| Design Arena: SVG | 1215 | 1240 | Grok 4.6 |
| Design Arena: ASCII art | 1256 | 1279 | Grok 4.6 |
| LiveBench design | 78.1% | 79.8% | Grok 4.6 |
GeneralLike-for-likePublic boards, shown as published, kept out of the index256
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Grok 4.5 | Grok 4.6 | 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
Same cost
Repository review
50K input + 3K output tokens
Same cost
Agent loop
200K input + 10K output tokens
Same cost
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 | SpaceXAI Grok 4.5 | SpaceXAI Grok 4.6 |
|---|---|---|
| Maker | xAI | xAI |
| Weights | No public weights listed | No public weights listed |
| Released | July 8, 2026 | August 12, 2026 |
| Context window | 500K | 500K |
| Input $/Mtok | $2.00 | $2.00 |
| Output $/Mtok | $6.00 | $6.00 |
| Coding index | 58.6 | 65.7 |
| Independent measurements behind it | 6 | 6 |
| Cost per solved task | $0.62 | $0.82 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Grok 4.5 better than Grok 4.6 for coding?
Pick Grok 4.6 for coding: it scores 65.7 on the Vibe Coding Index against 58.6 for Grok 4.5. Pick Grok 4.5 for a lower cost per solved task ($0.62 against $0.82) and work measured by LiveBench coding or SkillsBench (Vals AI), where it leads.
Which is cheaper, Grok 4.5 or Grok 4.6?
Grok 4.5 costs $2.00 per million input tokens and $6.00 per million output tokens; Grok 4.6 costs $2.00 and $6.00. Per solved SWE-bench Verified task, Grok 4.5 costs $0.62 and Grok 4.6 $0.82.
Which has the bigger context window, Grok 4.5 or Grok 4.6?
Neither: both take 500K tokens.
How many benchmarks is the Grok 4.5 vs Grok 4.6 comparison based on?
114 benchmarks have a result for both models as of October 22, 2026. Grok 4.5's coding index rests on 6 independent benchmarks and Grok 4.6's on 6. Where only one model has a result, the page says so instead of naming a winner.