XAI · MODEL TELEMETRY CARD

Grok 4.6

Canonical UID: xai/grok-4.6 · Released: 2026-08-12

Measured configurations of Grok 4.6 across benchmark suites. Exact machine values in site.json and data dumps.
Domain Benchmark Effort / Mode ⚡ Value/Cost # 🏆 Performance # Score Task $ Unit Cost $ Unit
Code ALE-Bench XHIGH #5 / 94 #21 / 94 1508.08 $0.24005 $0.0001592 rating point
Reason ProofBench STD #15 / 62 #27 / 62 0.51 $0.759043 $1.488 correct answer
Agent DeepSWE LOW #5 / 65 #52 / 65 0.416 $1.04236 $2.503 solved task
Agent DeepSWE MEDIUM #22 / 65 #19 / 65 0.675 $3.44898 $5.111 solved task
Agent DeepSWE HIGH #29 / 65 #27 / 65 0.652 $4.38487 $6.726 solved task
Agent DeepSWE XHIGH #35 / 65 #23 / 65 0.667 $5.49767 $8.237 solved task

Compare against all models on ← #frontier Live Ranking Board · Fixed-capability cost half-life analysis on /methodology.