Compare / head-to-head
Grok 4.6vs
Step 5 Preview
Grok 4.6 leads 3 of 5 shared benchmarks. Step 5 Preview is 2.2x cheaper per token. Step 5 Preview has the larger context window (1M tokens).
Benchmarks from cited public sources; pricing from official pages; status from official provider feeds.
Shared benchmarks
3 – 1
Grok 4.6 leads · 1 tied
Cheaper per token
Step 5 Preview
2.2x cheaper, input + output
Larger context
Step 5 Preview
1M tokens
Providers
2 providers
xAI · StepFun
Grok 4.6
xAI · released 2026-08-12
textvision
Step 5 Preview
StepFun · released 2026-09-20
textvisionvideo
Quality
Benchmark matrix
5 shared · 33 only Grok 4.6 · 0 only Step 5 Preview| Benchmark | Grok 4.6 | Step 5 Preview | Δ | Edge |
|---|---|---|---|---|
| Reported by both · 5 | ||||
| LMArena Elo | 1453.8 ↗ | 1453.7 ↗ | +0.1 | Grok 4.6 |
| LMArena Agent (LMArena) | 0.0128 ↗ | 0.0048 ↗ | 0 | Tie |
| LMArena Vision (LMArena) | 1263.5 ↗ | 1259.7 ↗ | +3.8 | Grok 4.6 |
| LMArena WebDev (LMArena) | 1619.5 ↗ | 1570.2 ↗ | +49.3 | Grok 4.6 |
| Terminal-Bench 4.0 (Vals AI) | 17.2% ↗ | 31.8% ↗ | -14.6 pt | Step 5 Preview |
| Only Grok 4.6 reports · 33 | ||||
| GPQA Diamond | 94% ↗epoch run | not reported | — | — |
| AA-Briefcase | 1577 ↗ | not reported | — | — |
| APEX-Agents | 57.5% ↗ | not reported | — | — |
| APEX-Agents (Mercor) | 65.3% ↗ | not reported | — | — |
| APEX-SWE | 56.4% ↗ | not reported | — | — |
| Artificial Analysis Intelligence Index | 61 ↗ | not reported | — | — |
| Chess Puzzles (Epoch AI run) | 40% ↗epoch run | not reported | — | — |
| CursorBench 3.2 | 69.9% ↗ | not reported | — | — |
| DeepSWE v1.1 | 65.9% ↗ | not reported | — | — |
| DeepSWE v1.1 (Datacurve) | 67.5% ↗ | not reported | — | — |
| EBR-bench (Epoch AI run) | 30.5% ↗epoch run | not reported | — | — |
| FrontierCode v1.1 (Extended) | 61.3% ↗ | not reported | — | — |
| FrontierMath Tier 4 v2 (Epoch AI run) | 31.7% ↗epoch run | not reported | — | — |
| FrontierMath Tiers 1-3 v2 (Epoch AI run) | 66% ↗epoch run | not reported | — | — |
| FrontierSWE V2 (Proximal Labs) | 25.3% ↗ | not reported | — | — |
| Furniture Assembly (Epoch AI run) | 40% ↗epoch run | not reported | — | — |
| GDPVal-AA v2 | 1753 ↗ | not reported | — | — |
| Harvey LAB (Vals) | 15.8% ↗ | not reported | — | — |
| LiveBench Agentic Coding (LiveBench) | 57% ↗ | not reported | — | — |
| LiveBench Coding (LiveBench) | 76.8% ↗ | not reported | — | — |
| LiveBench Data Analysis (LiveBench) | 73.9% ↗ | not reported | — | — |
| LiveBench Instruction Following (LiveBench) | 71.9% ↗ | not reported | — | — |
| LiveBench Language (LiveBench) | 83.7% ↗ | not reported | — | — |
| LiveBench Mathematics (LiveBench) | 92.6% ↗ | not reported | — | — |
| LiveBench Reasoning (LiveBench) | 90.5% ↗ | not reported | — | — |
| Mystery Game Puzzles (Epoch AI run) | 34% ↗epoch run | not reported | — | — |
| OTIS Mock AIME 2024-2025 (Epoch AI run) | 99.2% ↗epoch run | not reported | — | — |
| SAGE (Vals AI) | 28.9% ↗ | not reported | — | — |
| SimpleBench (SimpleBench) | 75.9% ↗ | not reported | — | — |
| SimpleQA Verified | 49.3% ↗epoch run | not reported | — | — |
| Terminal-Bench 3.0 | 26% ↗ | not reported | — | — |
| Vending-Bench 2 (Andon Labs) | 9047.03 ↗ | not reported | — | — |
| WeirdML (Håvard Tveit Ihle) | 67.3% ↗ | not reported | — | — |
| Only Step 5 Preview reports · 0 | ||||
| Step 5 Preview reports nothing Grok 4.6 does not. | ||||
Scores tagged "epoch" or "matharena" are independent runs, used only where the lab has not published its own; ⚠ marks rows MathArena flags as released after the competition. Higher is better on every row. Δ is Grok 4.6 minus Step 5 Preview in the benchmark's own unit. "Not reported" means the lab has not published that figure; it is not a zero. ↗ opens the source.
Specs & pricing
Side by side
Official pricing pages and model cards| Spec | Grok 4.6 | Step 5 Preview | Edge |
|---|---|---|---|
| Context window | 500K tokens | 1M tokens | Step 5 Preview |
| Max output | — | 64K tokens | — |
| Input price / 1M | $2 ↗ | $1 ↗ | Step 5 Preview |
| Output price / 1M | $6 ↗ | $2.7 ↗ | Step 5 Preview |
| Cached input / 1M | $0.5 ↗ | $0.05 ↗ | Step 5 Preview |
| Input + output / 1M Lower is cheaper. List prices; batch, tool and regional fees excluded. | $8 | $3.7 | Step 5 Preview |
| Modalities | text · vision | text · vision · video | Step 5 Preview |
| Released | 2026-08-12 | 2026-09-20 | — |
| Cited benchmark scores | 43 | 5 | — |
Reliability
Provider status
Live from /statusMore matchups
Grok 4.6 vs …
Models sharing the most benchmarksMore matchups
Step 5 Preview vs …
Models sharing the most benchmarksBuilt by Respan
Which one wins on your data?
Public benchmarks are a starting point. Run Grok 4.6 and Step 5 Preview on your own prompts with Respan evals, or route to either through one gateway key with automatic failover.