Grok 4.6
Grok 4.6 is an API-only (closed weights) language model from xAI, released August 12, 2026. Its makers disclosed about 1.5T parameters. The context window is 488K tokens. On the Epoch Capabilities Index it scores 157.8, rank 28 of 256 rated models. API price: $2 input / $6 output per million tokens.
Today we are releasing Grok 4.6. Grok 4.6 builds on Grok 4.5 with a particular focus on long-running agents and more ambitious interactive and visual work. It stays with complex tasks across many steps, whether researching a topic, analyzing information, working across a codebase, or turning an idea into a polished …
Check it against your own PC →
Strengths by category
ModelTape's 0–100 fit on Epoch AI benchmark results.
| Category | Score | Based on |
|---|---|---|
| Reasoning | 25.0 | Chess Puzzles, SimpleBench, Mystery Game Puzzles |
| Math | 53.7 | ProofBench, OTIS Mock AIME 2024-2025, FrontierMath-Tiers-1-3-v2-Private, FrontierMath-Tier-4-v2-Private |
| Coding | 59.0 | MirrorCode, FrontierSWE, DeepSWE |
| Knowledge | 77.2 | GPQA diamond, SimpleQA Verified |
Benchmark results
| Benchmark (Epoch AI) | Score |
|---|---|
| OTIS Mock AIME | 99.2% |
| GPQA Diamond | 95.8% |
| SimpleBench | 71.1% |
| FrontierMath (Tiers 1–3) | 69.4% |
| DeepSWE | 67.5% |
| ProofBench | 51.0% |
| SimpleQA Verified | 49.3% |
| MirrorCode | 41.1% |
| FrontierMath-Tier-4-v2-Private | 31.7% |
| EBR-bench | 30.5% |
| Chess Puzzles | 29.5% |
| Mystery Game Puzzles | 27.3% |
| FrontierSWE | 25.3% |
| Furniture Assembly | 14.3% |
Links
Announcement · Epoch AI benchmarks
Models with a similar score
Muse Wonder · Kimi K3.1 Pro · GPT-5.6 Terra · Claude Haiku 5.5 · GPT-5.4 Pro · GPT-5.5 · Claude Opus 4.8 · DeepSeek V4.1 Flash 1017
More from xAI
Grok 4.7 · Grok 4.5 · Grok 4.3 Beta · Grok Build 0.1 · Grok 4.20 Multi-Agent · Grok 4.20 · Grok 4.1 Fast · Grok 4.1