Claude Sonnet 5.5 vs Grok 4.3: pricing compared
For a 30-step coding agent run, Grok 4.3 is 1.3x cheaper ($0.64 vs $0.83). Prices verified 2026-10-18.
| Claude Sonnet 5.5 | Grok 4.3 | |
|---|---|---|
| Provider | Anthropic | xAI |
| Input / 1M | $2.00 | $1.25 |
| Cached input / 1M | $0.10 | $0.20 |
| Output / 1M | $10.00 | $2.50 |
| Batch discount | 50% | 20% |
| Context window | 1M | 1M |
| Long-context tier | none | above 200K |
Cost on real workloads
| Workload | Claude Sonnet 5.5 | Grok 4.3 |
|---|---|---|
| 10,000 chatbot replies 2,000 input + 400 output tokens each, 50% of the prompt cached | $61.00 | $24.50 |
| One 30-step coding agent run 12K system + tools prompt, 1K output and 3K tool result per step, caching on | $0.83 | $0.64 |
| 1M documents classified (batch) 600 input + 20 output tokens each, batch API where available | $700 | $640 |
Price is only half the decision: a cheaper model that needs more steps or retries can cost more per finished task. Run your own numbers in the calculator (both models pinned), or see how both rank in the use-case guides.
FAQ
Is Claude Sonnet 5.5 cheaper than Grok 4.3?
For a typical 30-step coding agent run with caching, Grok 4.3 costs $0.64 and Claude Sonnet 5.5 costs $0.83, so Grok 4.3 is 1.3x cheaper. Per 1M tokens, Claude Sonnet 5.5 is $2.00 / $10.00 and Grok 4.3 is $1.25 / $2.50.
Which has the bigger context window?
Claude Sonnet 5.5: 1,000,000 tokens. Grok 4.3: 1,000,000 tokens.