Noometry

xAI, proprietary

Grok 4 Fast

Grok 4 Fast by xAI ranks 167th of 354 ranked models on the Noometry Index as of October 2026, with a score of 39.4. Its strongest category is long context, where it ranks 3rd.

Last verified

Noometry rank
#167 of 354
Index score
39.4
Evidence
Confirmed 30 results
Provider
xAI
Released
September 19, 2025
Weights
Proprietary
Reasoning
Unknown
Context window
—
Max output
—
Input price
Not listed
Output price
Not listed
Blended price
Not listed
Output speed
577 tokens/s Kagi
Value
Not ranked
Knowledge cutoff
Unknown

Category scores

Each category score combines every public result we have in that category.

  1. Coding 32.5
  2. Agentic & Tool Use 29.5
  3. Reasoning 22.2
  4. Math 38.9
  5. Knowledge 32.0
  6. Multilingual 51.3
  7. Instruction Following 73.2
  8. Long Context 63.2
  9. Writing & Preference 60.0
CategoryScoreRankResults
Coding32.5#2713
Agentic & Tool Use29.5#862
Reasoning22.2#2015
Math38.9#1231
Knowledge32.0#2142
Multilingual51.3#1111
Instruction Following73.2#1211
Long Context63.2#32
Writing & Preference60.0#1023

Strengths and weaknesses

Categories where Grok 4 Fast places highest and lowest among the models ranked in each, with its score against that category's median.

Strongest categories

CategoryScorevs medianRank
Long Context63.2+22.3#3 of 296, top 2%
Writing & Preference60.0+6.2#102 of 312, top 33%
Multilingual51.3+3.9#111 of 297, top 38%

Weakest categories

CategoryScorevs medianRank
Coding32.5−6.3#271 of 340, top 80%
Knowledge32.0−5.3#214 of 314, top 69%
Reasoning22.2−1.4#201 of 350, top 58%

Closest competitors

The models ranked just above and below Grok 4 Fast. When scores are this close, price and speed are often the better way to choose.

ModelRankScoreBlended $/MSpeed
Hunyuan Turbo 0110#16339.6——Compare
Claude 3.7 Sonnet#16439.5——Compare
Claude Haiku 4.5#16539.5$2—Compare
DeepSeek-V3#16639.5$0.4173Compare
Olmo 3.1 32b Instruct#16839.4——Compare
Granite 4.2 3b#16939.4——Compare
Gemini 2.5 Flash#17039.3$0.85152Compare
Step 2 16k Exp 202412#17139.2——Compare

Sponsored placements are available on pages like this one. Advertise on Noometry

Benchmark results

Every published result we track, with its source. Bold rows are the ones used for ranking; where several exist we prefer independent runs over self-reported numbers.

Coding

BenchmarkScorePositionSettingSourceDate
LMArena WebDev1159#112 of 113, top 100%LMArena2026-10-08
WeirdML42.9%#67 of 119, top 57%Epoch AI
LMArena Coding1429#111 of 294, top 38%LMArena2026-10-08
LMArena Coding1420LMArena2026-10-08

Agentic & Tool Use

BenchmarkScorePositionSettingSourceDate
τ²-bench Banking15.7%#22 of 26, top 85%highτ²-bench2026-05-05
Cybench30%#9 of 21, top 43%Epoch AI
LMArena Search1171#21 of 32, top 66%LMArena2026-08-24

Reasoning

BenchmarkScorePositionSettingSourceDate
ARC-AGI-25.3%#55 of 83, top 67%Epoch AI
Kagi LLM Benchmark66.1%#32 of 99, top 33%Kagi LLM Benchmark
Kagi LLM Benchmark35.4%Kagi LLM Benchmark
ARC-AGI-148.5%#57 of 83, top 69%Epoch AI
LMArena Hard Prompts1405LMArena2026-10-08
LMArena Hard Prompts1412#115 of 297, top 39%LMArena2026-10-08
DTBench82.7%#63 of 151, top 42%Epoch AI
Epoch Capabilities Index144.2#86 of 213, top 41%Epoch AI2025-09-19
ForecastBench60.5#26 of 72, top 37%Epoch AI

Math

BenchmarkScorePositionSettingSourceDate
LMArena Math1403LMArena2026-10-08
LMArena Math1419#103 of 285, top 37%LMArena2026-10-08

Knowledge

BenchmarkScorePositionSettingSourceDate
Vectara Hallucination Rate (lower is better)20.2%Vectara Hallucination Leaderboard
Vectara Hallucination Rate (lower is better)19.7%#92 of 96, top 96%Vectara Hallucination Leaderboard
LMArena Expert1399LMArena2026-10-08
LMArena Expert1411#112 of 273, top 42%LMArena2026-10-08

Multilingual

BenchmarkScorePositionSettingSourceDate
LMArena Non-English1396#111 of 297, top 38%LMArena2026-10-08
LMArena Non-English1389LMArena2026-10-08
LMArena Chinese1457#94 of 285, top 33%LMArena2026-10-08
LMArena Chinese1436LMArena2026-10-08
LMArena French1431#90 of 223, top 41%LMArena2026-10-08
LMArena German1383#108 of 231, top 47%LMArena2026-10-08
LMArena Japanese1352#95 of 211, top 46%LMArena2026-10-08
LMArena Korean1332LMArena2026-10-08
LMArena Korean1357#99 of 213, top 47%LMArena2026-10-08
LMArena Russian1389#120 of 283, top 43%LMArena2026-10-08
LMArena Russian1384LMArena2026-10-08
LMArena Spanish1415#98 of 226, top 44%LMArena2026-10-08
LMArena Spanish1404LMArena2026-10-08

Instruction Following

BenchmarkScorePositionSettingSourceDate
LMArena Instruction Following1387#115 of 298, top 39%LMArena2026-10-08
LMArena Instruction Following1383LMArena2026-10-08

Long Context

BenchmarkScorePositionSettingSourceDate
Fiction.LiveBench94.4%#4 of 47, top 9%Epoch AI
LMArena Longer Query1415#99 of 291, top 35%LMArena2026-10-08
LMArena Longer Query1412LMArena2026-10-08

Writing & Preference

BenchmarkScorePositionSettingSourceDate
LMArena Text1407#113 of 297, top 39%LMArena2026-10-08
LMArena Text1400LMArena2026-10-08
LMArena Creative Writing1378LMArena2026-10-08
LMArena Creative Writing1387#98 of 295, top 34%LMArena2026-10-08
LMArena Multi-Turn1414#104 of 295, top 36%LMArena2026-10-08
LMArena Multi-Turn1404LMArena2026-10-08

Compare Grok 4 Fast

Other xAI models

Frequently asked questions

How good is Grok 4 Fast?

Grok 4 Fast by xAI ranks 167th of 354 ranked models on the Noometry Index as of October 2026, with a score of 39.4. Its strongest category is long context, where it ranks 3rd.

Is Grok 4 Fast open source?

No. Grok 4 Fast is proprietary and available only through xAI's API and partner platforms.

How fast is Grok 4 Fast?

Grok 4 Fast generated about 577 output tokens per second in the Kagi LLM Benchmark's timed runs. Speed varies by provider, load and reasoning effort.

What are Grok 4 Fast's strengths and weaknesses?

Relative to other ranked models, Grok 4 Fast places best in long context, writing & preference, multilingual and lowest in coding, knowledge, reasoning.

What is Grok 4 Fast best at?

Its best category is long context, where it ranks 3rd on Noometry.