token.app › Compare › Qwen3 235B A22B Thinking 2507 vs Grok 4.6

Qwen3 235B A22B Thinking 2507 vs Grok 4.6

Qwen3 235B A22B Thinking 2507 is 2.5× cheaper per token on a blended 3:1 basis. Grok 4.6 accepts the larger context window (500K).

Head to head
Blended price uses a 3:1 input:output mix
Qwen3 235B A22B Thinking 2507Grok 4.6
ProviderAlibabaSpaceXAI
Input $/1M$0.450$2.00
Output $/1M$3.50$6.00
Blended $/1M$1.21$3.00
Context window128K500K
Max output16K450K
ReleasedJul 2025Aug 2026
LicenceOpen weightsProprietary
Choose Qwen3 235B A22B Thinking 2507 if…
  • Costs less per token — $1.21 vs $3.00 blended
  • Open weights — self-hostable
Choose Grok 4.6 if…
  • Larger context window (500K)
  • Newer model (released Aug 2026)
FAQ
Which is better, Qwen3 235B A22B Thinking 2507 or Grok 4.6?
We do not hold independently-run benchmark scores covering both Qwen3 235B A22B Thinking 2507 and Grok 4.6, so we make no quality claim. Their pricing and specifications are compared on this page.
Is Qwen3 235B A22B Thinking 2507 cheaper than Grok 4.6?
Qwen3 235B A22B Thinking 2507 is cheaper: $1.21 versus $3.00 per million tokens blended at 3:1 input:output. Input alone: $0.450 vs $2.00. Output alone: $3.50 vs $6.00.
What context windows do Qwen3 235B A22B Thinking 2507 and Grok 4.6 have?
Qwen3 235B A22B Thinking 2507: 128,000 tokens. Grok 4.6: 500,000 tokens.
Who makes Qwen3 235B A22B Thinking 2507 and Grok 4.6?
Qwen3 235B A22B Thinking 2507 is made by Alibaba. Grok 4.6 is made by SpaceXAI.
Related comparisons
Qwen3 235B A22B Thinking 2507 full details Grok 4.6 full details Qwen3 235B A22B Thinking 2507 vs Flint (free) Qwen3 235B A22B Thinking 2507 vs MiniMax M3.1 Flash (free) Qwen3 235B A22B Thinking 2507 vs DeepSeek V4.1 Flash 1017 Qwen3 235B A22B Thinking 2507 vs Glimpse Qwen3 235B A22B Thinking 2507 vs Xing4.0 29B A4B (free) Qwen3 235B A22B Thinking 2507 vs Beam 1.0 501B A23B