Grok 4.7 vs Qwen3.8-Max.
Verdict by category, tale of the tape, shared benchmarks only, and cost at three workloads — all from the tracked data.
Verdict by category
| Category | Grok 4.7 | Qwen3.8-Max | Winner |
|---|---|---|---|
| League Score | 66.8 | 65.0 | Grok 4.7 |
| Coding (SWE-bench %) | — | — | — |
| Reasoning (GPQA %) | — | 92.6 | — |
| Context (tokens) | 500000 | 1000000 | Qwen3.8-Max |
| Speed (tok/s) | 87.6 | 37 | Grok 4.7 |
| Price (blended $/1M) | $3.00 | $3.00 | Qwen3.8-Max (cheaper) |
Tale of the tape
| Spec | Grok 4.7 | Qwen3.8-Max |
|---|---|---|
| Maker | xAI (SpaceXAI) | Alibaba Qwen |
| Released | Sep 21, 2026 | Aug 3, 2026 |
| League Score | 67 | 65 |
| Context | 500,000 tokens | 1,000,000 tokens |
| Max output | — | 131,072 |
| $/1M input | $2.00 | $2.00 |
| $/1M output | $6.00 | $6.00 |
| Weights | Closed | Open |
| Speed | 87.6 t/s | 37 t/s |
Head-to-head benchmarks (shared only)
| Benchmark | Grok 4.7 | Qwen3.8-Max |
|---|---|---|
| LiveBench overall | 70.1 | 61.6 |
Cost at three workloads
| Workload | Grok 4.7 | Qwen3.8-Max |
|---|---|---|
| Light — 1M in / 250K out | $3.50 | $3.50 |
| Standard — 10M in / 2.5M out | $35.00 | $35.00 |
| Long-context — 50M in / 5M out | $130.00 | $130.00 |
The pick
Pick Grok 4.7 if you want the higher overall League Score.
Pick Qwen3.8-Max if you need the bigger context window; you want downloadable weights / self-hosting.
All figures from the model database — sources and dates on each model page. Cost your own workload or compare a different pair.
Frequently asked questions
Which is better overall, Grok 4.7 or Qwen3.8-Max?
Grok 4.7 leads on overall League Score, 67 to 65 — a gap of 2 points on a composite of sourced benchmarks. That is overall form, not every category: check the per-category verdicts before choosing.
Which is cheaper, Grok 4.7 or Qwen3.8-Max?
They cost the same on our blended measure: $3.00 per 1M tokens each.
Which has the bigger context window, Grok 4.7 or Qwen3.8-Max?
Qwen3.8-Max, with 1M of context against 500K for Grok 4.7.
Are Grok 4.7 and Qwen3.8-Max open source?
Grok 4.7 is closed weights — API access only; Qwen3.8-Max is open weights — it can be downloaded and self-hosted.