Grok 4.6vsGrok 4.5
Across 10 shared benchmarks, Grok 4.6 leads overall: Grok 4.6 wins 10, Grok 4.5 wins 0, with 0 ties and an average score difference of +54.32.
Grok 4.610 wins(100%)(0%)0 winsGrok 4.5
Benchmark scores
Grouped by capability, sorted by largest gap within each. 10 shared benchmarks.
Coding and Software Engineer
Grok 4.6 4/4| Benchmark | Grok 4.6 | Grok 4.5 | Diff |
|---|---|---|---|
| DeepSWE | 65.909 / 26Thinking High (With Tools) | 5317 / 26Thinking High (With Tools) | +12.90 |
| FrontierCode 1.1 | 61.302 / 5Thinking High (With Tools) | 56.604 / 5Thinking High (With Tools) | +4.70 |
| CursorBench 3.2 | 69.902 / 4Thinking High (With Tools) | 66.704 / 4Thinking High (With Tools) | +3.20 |
| APEX-SWE | 56.402 / 3Thinking High (With Tools) | 53.603 / 3Thinking High (With Tools) | +2.80 |
Productivity Knowledge
Grok 4.6 3/3| Benchmark | Grok 4.6 | Grok 4.5 | Diff |
|---|---|---|---|
| AA-Briefcase | 1,5772 / 6Thinking High (With Tools) | 1,3136 / 6Thinking High (With Tools) | +264 |
| GDPval-AA v2 | 1,7533 / 13Thinking High (With Tools) | 1,5267 / 13Thinking High (With Tools) | +227 |
| Harvey Lab-AA | 15.803 / 6Thinking High (With Tools) | 12.904 / 6Thinking High (With Tools) | +2.90 |
Agent Level Benchmark
Grok 4.6 1/1| Benchmark | Grok 4.6 | Grok 4.5 | Diff |
|---|---|---|---|
| APEX-Agents | 57.502 / 5Thinking High (With Tools) | 47.104 / 5Thinking High (With Tools) | +10.40 |
AI Agent - Tool Usage
Grok 4.6 1/1| Benchmark | Grok 4.6 | Grok 4.5 | Diff |
|---|---|---|---|
| Terminal-Bench 3.0 | 264 / 6Thinking High (With Tools) | 15.705 / 6Thinking High (With Tools) | +10.30 |
General Knowledge
Grok 4.6 1/1| Benchmark | Grok 4.6 | Grok 4.5 | Diff |
|---|---|---|---|
| AA Intelligence Index | 612 / 8Thinking High (With Tools) | 564 / 8Thinking High (With Tools) | +5 |
Specs
| Field | Grok 4.6 | Grok 4.5 |
|---|---|---|
| Publisher | xAI | xAI |
| Release date | 2026-08-12 | 2026-07-08 |
| Model type | Coding model | Coding model |
| Architecture | Dense | Dense |
| Parameters | Not available | Not available |
| Context length | 500K | 500K |
| Max output | Not available | Not available |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Grok 4.6 | Grok 4.5 |
|---|---|---|
| Text input | $2 / 1M tokens | $2 / 1M tokens |
| Text output | $6 / 1M tokens | $6 / 1M tokens |
| Cache read | $0.5 / 1M tokens | $0.5 / 1M tokens |
Summary
- Grok 4.6leads in:Coding and Software Engineer (4/4), Productivity Knowledge (3/3), Agent Level Benchmark (1/1), AI Agent - Tool Usage (1/1), General Knowledge (1/1)
On average across the 10 shared benchmarks, Grok 4.6 scores 54.32 higher.
Largest single-benchmark gap: AA-Briefcase — Grok 4.6 1,577 vs Grok 4.5 1,313 (+264).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.