Grok 4.6
xAI
65.3#60 of 427 overall
- Input / output
- US$ 2,00 / US$ 6,00 US$ per 1M tokens
- Context
- 500K tokens
- Release
- 08/12/2026 release date
- Weights
- Closed
- Speed
- 0 tok/s first token in 0.00 s
Family history
The 11 versions of this family, from first to newest, by release date. Click a point to open that version.
See the data as a table
| Model | Released | Score | Rank | AA Intel. | Elo Arena |
|---|---|---|---|---|---|
| Grok 2 | 12/12/2024 | 20.5 | #401 | — | 1305 |
| Grok 3 | 02/19/2025 | 45.6 | #192 | — | 1425 |
| Grok 4 | 07/10/2025 | 54.0 | #111 | — | 1411 |
| Grok 4.1 | — | — | — | — | 1436 |
| Grok 4.20 | 03/31/2026 | 62.7 | #71 | — | 1451 |
| Grok 4.20 0309 v2 | 04/07/2026 | 66.8 | #55 | — | — |
| Grok 4.3 | 04/30/2026 | 44.0 | #206 | — | 1397 |
| Grok Latest | 07/08/2026 | 34.5 | #308 | — | — |
| Grok 4.5 | 07/08/2026 | 72.9 | #32 | — | 1448 |
| Grok 4.6 | 08/12/2026 | 65.3 | #60 | — | 1429 |
| Grok 4.7 | 09/21/2026 | 66.4 | #57 | 35.0 | 1398 |
Scores by source
Scores by sourcepercentile = % of ranked models it beats| Metric | Value | Percentile | Source |
|---|---|---|---|
| LiveBench weight 3 | 67.8 | 55% | LiveBench |
| Arena text (Elo) weight 2 | 1429 | 62% | LMArena |
| Arena WebDev (Elo) weight 1 | 1616 | 80% | LMArena |
| Arena vision (Elo) weight 0.5 | 1263 | 61% | LMArena |
| Humanity's Last Exam weight 1 | 44.1 | 94% | Artificial Analysis |
| Terminal-Bench weight 1 | 17.2 | 78% | Artificial Analysis |
| GPQA Diamond | 93.5 | 98% | Artificial Analysis |
| Long-context reasoning | 81.0 | 90% | Artificial Analysis |
| LiveBench long agentic | 28.0 | 45% | LiveBench |
| LiveBench reasoning | 91.2 | 81% | LiveBench |
| LiveBench coding | 89.1 | 65% | LiveBench |
| LiveBench browser | 49.4 | 44% | LiveBench |
| LiveBench data analysis | 75.3 | 42% | LiveBench |
| LiveBench research | 84.5 | 87% | LiveBench |
| LiveBench design | 79.8 | 71% | LiveBench |
By category
| Category | Score | Rank | Metrics |
|---|---|---|---|
| Agentic | 57.6 | #42 | LiveBench long agentic, Terminal-Bench |
| Coding | 64.7 | #21 | LiveBench coding, Arena WebDev (Elo) |
| Reasoning | 80.7 | #10 | LiveBench reasoning, Humanity's Last Exam, GPQA Diamond |
| Data analysis | 46.0 | #19 | LiveBench data analysis |
| Research | 75.7 | #5 | LiveBench research, Long-context reasoning |
| Design | 66.9 | #17 | LiveBench design, Arena WebDev (Elo) |
| Browser use | 46.8 | #19 | LiveBench browser |
| Chat | 56.0 | #75 | Arena text (Elo) |
| Vision | 55.5 | #42 | Arena vision (Elo) |
History
RankingIA score on each collection day (hover to see the rank).
Daily collection started on 10/30/2026. The chart appears from the second day the model is ranked.