AI Agent StoreLow
Tested setting-9.2 pts vs High
$0.4753 / source task
10,264 output tokens
xAI · REASONING EFFORT GUIDEGrok 4.6
Grok 4.6
How much thinking?
Compare Low vs Medium vs High vs Extra high where tested. Find the effort setting that fits your task, and see how much extra reasoning changes the result.
SAME MODEL. DIFFERENT THINKING.
Choose a thinking level for Grok 4.6
Artificial Analysis
Artificial Analysis composite of ten evaluations, on a 0–100 scale. Index points are not percent correct. The same index version is used for every effort; costs and tokens reflect the publisher’s weighted task mix.
Grok 4.6 API · Intelligence Index v4.3.2 · four effort settings · Snapshot reviewed: 2026-09-23 · Reviewed 2026-10-15Check the source ↗SUGGESTED SETTING TO TESTHigh
Lowest measured cost within 1.0 points of the best tested intelligence index.
44.3 / 100 · $1.86 / source taskExact best1.0 pts10 pts
Set 0 for the highest score. Increase the gap to explore cheaper or lighter options.
THE EFFORT CURVE
Does more thinking help?
● Near-best pickOrange = score decreased
Scores out of 100 · Full 0–100 scale · tap a point to select mode B. Cyan band: within 1.0 points of the best tested score.
WHAT YOU SPEND
Cost of more effort
Measured average USD per benchmark task.
Low$0.4753
Medium$1.5
HighNear-best pick$1.86
Extra high$2.32
Measured in this study with its particular tools and setup. Your task mix, tools, caching and endpoint can change the result.
Score difference · B minus A-0.1 pts
Small gap · B scored lower
Measured cost · B / A1.25×
Same benchmark workload
Highest observed score44.3
High
We describe gaps under 1 point as small, 1–under 5 as noticeable, and 5 or more as large. These are reading aids, not proof that a difference would hold up in repeated tests. Equal scores can still hide different mistakes.
When more thinking scored lower
High → Extra high-0.1 pts44.3 → 44.2 on intelligence index
EVERY TESTED MODE
4 modes · untested settings are left blankResults for each tested setting
35.1 / 100
Medium
Tested setting-1.5 pts vs High
$1.5 / source task
28,466 output tokens
High
Suggested setting0.0 pts vs High
$1.86 / source task
35,835 output tokens
Extra high
A better tradeoff is available-0.1 pts vs High
$2.32 / source task
37,608 output tokens
Full benchmark table & exact configuration IDs
Swipe inside the table for all measures. Figures from different study selections are not directly comparable.
| Effort | Intelligence Index | Cost / task | Median seconds | Output tokens | Exact source ID |
|---|---|---|---|---|---|
| Low | 35.1 | $0.4753 | — | 10,264 | grok-4-6-low |
| Medium | 42.8 | $1.5 | — | 28,466 | grok-4-6-medium |
| High | 44.3 | $1.86 | — | 35,835 | grok-4-6 |
| Extra high | 44.2 | $2.32 | — | 37,608 | grok-4-6-xhigh |
SCALE THE OBSERVED COST
Effort cost difference calculator
What would A versus B cost if you repeated this benchmark’s average workload?
A · High$1,858.94
B · Extra high$2,323.69
Difference$464.75more with B
A source-workload scenario using average measured USD costs. Source prices, caching and tools are already reflected in those observed costs. Your agent’s costs may differ.
YOUR WORKLOAD
Reasoning-token cost calculator
Estimate the extra bill from thinking tokens. The starting token counts are editable examples, not measurements of Grok 4.6. Rates start from the current model snapshot.
A · High$32estimated monthly cost
B · Extra high$56estimated monthly cost
Mode B costs $24 more per month with these assumptions.
This estimate uses the token counts and prices you enter. Excludes retries, tools, caching, long-context premiums, hosting and taxes. Includes thinking in output billing; do not also include it in “visible output.” Check endpoint limits separately. Changing mode does not automatically predict tokens or quality.
Grok 4.6 effort benchmarks at a glance
Each card compares settings tested together in one study. Scores from different benchmarks or question sets measure different work and are not averaged together.
General capability · Artificial Analysis v4.3.2
Artificial Analysis composite of ten evaluations, on a 0–100 scale. Index points are not percent correct. The same index version is used for every effort; costs and tokens reflect the publisher’s weighted task mix.
- Low
- 35.1 · $0.4753 / task
- Medium
- 42.8 · $1.5 / task
- High · pick
- 44.3 · $1.86 / task
- Extra high
- 44.2 · $2.32 / task
Near-best pick: High within 1 point of the best score. Highest-to-lowest score gap: 9.2 points.
Explore this study ↗All results and the exact settings tested
A dash means the study did not report that result. This table compares only the versions and settings tested on the same questions in this study.
| Effort / budget | Intelligence Index | Measured USD / task | Average output tokens | Exact source configuration |
|---|---|---|---|---|
| Low | 35.1 | $0.4753 | 10,264 | grok-4-6-low |
| Medium | 42.8 | $1.5 | 28,466 | grok-4-6-medium |
| High | 44.3 | $1.86 | 35,835 | grok-4-6 |
| Extra high | 44.2 | $2.32 | 37,608 | grok-4-6-xhigh |
Reasoning effort, explained
Which Grok 4.6 thinking level is most efficient?
High is the lowest-cost option within 1 point of the best intelligence index in Artificial Analysis. Extra high trails the best tested score by 0.1 points. This allows a score up to 1 point below the best result in the first listed study. Change the study, task and allowed score gap above to explore another comparison. It is a shortlist for testing your own workload.
Which Grok 4.6 effort settings are compared?
Low, Medium, High, Extra high appear in the collected studies. Each chart only includes modes tested together in that study. These published labels do not establish that every endpoint or application exposes every setting.
Does more effort improve Grok 4.6?
In Artificial Analysis, High to Extra high lowers the observed score by 0.1 points. The overall tested range is 9.2 points. These are observed differences, not proof that extra thinking caused the decline.
How much more does high effort cost for Grok 4.6?
Use the A/B selector and measured-cost calculator when the source reports dollar costs. For your own agent, enter average input, visible output and thinking tokens in the token calculator. Higher effort is adaptive, so the label itself cannot tell you the bill. Missing costs stay missing.