xAIproprietary
Released: 2026-06-15Grok 4
xAI fourth-generation foundation model trained on the expanded Colossus 200k H100/B200 cluster.
Intelligence Index
93.5/ 100
Calibrated multi-domain compositeThroughput Speed
75tok/sec
Output streaming throughputTime To First Token
420ms
Initial chunk server latencyLiveBench Contam-Free
81.2%
September 2026 suite scoreLiveBench Multi-Domain Evaluation Breakdown
Monthly refreshed contamination-free test setsLogical Reasoning94.5%
Coding & Repo Repair84.0%
Mathematics (AIME)95.0%
Data Analysis & Tables87.0%
Language & Comprehension89.5%
Instruction Following94.0%
Sourced Benchmark Evaluations (5)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| ARC-AGI-2 | Nearing Saturation | 73.7% | 75 | 2026-07-13 | independent | Source ↗ |
| SWE-bench Verified | Saturated | 63.8% | 25 | 2026-06-24 | independent | Source ↗ |
| IFEval | Deprecated | 94% | 25 | 2026-06-19 | independent | Source ↗ |
| GPQA | Nearing Saturation | 87% | 75 | 2024-06-01 | independent | Source ↗ |
| Humanity's Last Exam | Active | 24.5% | 90 | 2024-06-01 | independent | Source ↗ |
OpenAI Compatible API ExecutionAPI Ready
from openai import OpenAI
client = OpenAI()
response = client.chat.completions.create(
model="grok-4",
messages=[{"role": "user", "content": "Evaluate multi-step logic problem"}],
temperature=0.2
)
print(response.choices[0].message.content)Evaluation & Sourcing Notes
Scores listed for Grok 4 represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.
All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.