Claude Sonnet 5 vs Grok 4.3: pricing and benchmarks
Grok 4.3 is 38% cheaper on input tokens ($1.25 vs $2.00 per 1M). Grok 4.3 is 75% cheaper on output tokens ($2.50 vs $10.00 per 1M). On our document Q&A workload (50,000 questions a month), Grok 4.3 costs $575.00 a month versus $1,100 for Claude Sonnet 5, 48% less. Epoch AI has not published a capability index score for Grok 4.3 yet.
Prices verified September 14, 2026.
Which should you choose?
Epoch AI has not published a capability index score for Grok 4.3 yet.
Grok 4.3 costs $1.563 per 1M tokens (3:1 input-to-output mix) vs $4.00 for Claude Sonnet 5, 61% less.
Pricing and specs
| Claude Sonnet 5 | Grok 4.3 | |
|---|---|---|
| Provider | Anthropic | xAI |
| API model ID | claude-sonnet-5 | grok-4.3 |
| Input / 1M tokens | $2.00 | $1.25 |
| Cached input / 1M tokens | $0.20 | $0.20 |
| Output / 1M tokens | $10.00 | $2.50 |
| Long-context rates | — | $2.50 in / $5.00 out above 200K prompt tokens |
| Context window | 1M tokens | 1M tokens |
| Max output | 128K tokens | — |
| Batch discount | 50% off | — |
Benchmarks
| Claude Sonnet 5 | Grok 4.3 | |
|---|---|---|
| Epoch Capabilities Index | 156.2 | — |
| GPQA Diamond· Science reasoning | 87.4% | — |
| FrontierMath (Tiers 1–3)· Advanced math | 65.6% | — |
| SimpleQA Verified· Factual accuracy | 33.7% | — |
| ARC-AGI-2· Abstract reasoning | — | — |
| DeepSWE· Coding | 53.8% | — |
| APEX-Agents· Agentic work | 32.5% | — |
Source: Epoch AI, best recorded result per model (CC BY 4.0). A dash means no published score. See all benchmarks
Monthly cost for real workloads
| Workload | Claude Sonnet 5 | Grok 4.3 | Difference |
|---|---|---|---|
Support chatbot 100,000 replies a month, ~1,500 input tokens (system prompt + history) and ~400 output tokens each. | $700.00 | $287.50 | Grok 4.3 59% cheaper |
Document Q&A (RAG) 50,000 questions a month with ~8,000 tokens of retrieved context and ~600 output tokens each. | $1,100 | $575.00 | Grok 4.3 48% cheaper |
Coding agent 10,000 agent steps a month, ~40,000 input tokens each (70% read from the prompt cache) and ~2,000 output tokens. | $496.00 | $256.00 | Grok 4.3 48% cheaper |
Estimates use standard list prices and ignore cache-write surcharges. Different tokenizers can count the same text differently, so test with your own prompts.
Claude Sonnet 5 vs Grok 4.3 cost calculator
- $575.00/mo$0.01 / request
- Claude Sonnet 5Anthropic$1,100/mo$0.02 / request
Standard-tier list prices. Long-context rates apply automatically where the provider publishes a threshold. Excludes cache-write surcharges, taxes, batch and volume discounts.
Claude Sonnet 5 pricing notes
Anthropic’s best combination of speed and intelligence.
- The $2 / $10 launch price is now the standard price; the planned rise to $3 / $15 will not happen.
- Cache writes cost $2.50 (5-minute cache) or $4 (1-hour cache) per 1M tokens.
- Uses Anthropic’s newer tokenizer, which produces roughly 30% more tokens for the same text.
Grok 4.3 pricing notes
An earlier Grok 4 model with a 1M-token context window.
- Requests with 200K or more prompt tokens are billed at $2.50 input / $0.40 cached / $5 output for all tokens.
Frequently asked questions
- Is Claude Sonnet 5 cheaper than Grok 4.3?
- Grok 4.3 is 38% cheaper on input tokens ($1.25 vs $2.00 per 1M). Grok 4.3 is 75% cheaper on output tokens ($2.50 vs $10.00 per 1M). On our document Q&A workload (50,000 questions a month), Grok 4.3 costs $575.00 a month versus $1,100 for Claude Sonnet 5, 48% less.
- Which is more capable, Claude Sonnet 5 or Grok 4.3?
- Epoch AI has not published a capability index score for Grok 4.3 yet.
- How much does Claude Sonnet 5 cost per 1M tokens?
- Claude Sonnet 5 costs $2.00 per 1M input tokens and $10.00 per 1M output tokens, and $0.20 per 1M cached input tokens on Anthropic’s standard tier.
- How much does Grok 4.3 cost per 1M tokens?
- Grok 4.3 costs $1.25 per 1M input tokens and $2.50 per 1M output tokens, and $0.20 per 1M cached input tokens on xAI’s standard tier.
- Which has the larger context window, Claude Sonnet 5 or Grok 4.3?
- Both offer a 1M-token context window.