Solar System
loads default
Overall93
Score8/8
Tech97
UX / UI94 / 60
Time16 min 01 s
Cost$0.33
Compare →
Grok 4.6
xAIReleased 12/08/2026
Overall88/100#4 of 27
Tech86/100#17 of 27
UX / UI81 / 74#3 of 27
Score37.5/40
Total cost$1.83
Average time15 min 46 s#24 of 27
Total tokens309,323
API price: $2 / $6 per 1M input / output tokens · as of 18/10/2026 source ↗
What is Grok 4.6 good for?
xAI's flagship model focused on reasoning and agentic work (software engineering, STEM, knowledge work), with a 500K-token context.
Description based on the maker · source ↗
- Maker
- xAI
- Released
- 12/08/2026
- Context window
- 500K tokens
How did Grok 4.6 do in the AI Lab tests?
- 5 of 5 outputs load without errors.
- Its best technical result was on the “Solar System” prompt (97/100).
- It implemented the chess rules flawlessly – passing the objective perft test on all 6 positions.
- Average time per prompt 15 min 46 s#24 of 27
- Average cost per prompt $0.37#21 of 27
- Overall score 88/100#4 of 27
- Technical health 86/100#17 of 27
- Automatic UX rating 81/100#3 of 27
Outputs
All outputs of Grok 4.6 on the 5 prompts every other model got as well.
GTA-style game
loads default
Overall89
Score7.5/8
Tech57
UX / UI100 / 80
Time19 min 19 s
Cost$0.46
Compare →
Monitor test
loads default
Overall93
Score8/8
Tech97
UX / UI73 / 80
Time11 min 20 s
Cost$0.29
Compare →
New York & Central Park map
loads default
Overall79
Score6.5/8
Tech81
UX / UI67 / 80
Time18 min 21 s
Cost$0.43
Compare →
Chess with an engine
loads default
Overall87
Score7.5/8
Tech97
UX / UI73 / 70
Time13 min 51 s
Cost$0.32
Compare →
Cost = tokens priced with the API price list valid on the date shown next to the cost (prices change). ≈ = estimate (only the total token count is known), no ≈ = exact from input / output tokens. “Loads” = the page opens without errors; technical health also checks behaviour during use – judge the quality in the viewer.