Anthropic: Claude 3.5 Haiku vs xAI: Grok 4
Head-to-head API pricing and cost comparison between Anthropic’s Anthropic: Claude 3.5 Haiku and xAI’s xAI: Grok 4. Prices auto-refresh daily from OpenRouter.
Anthropic: Claude 3.5 Haiku is 73% cheaper for input tokens; Anthropic: Claude 3.5 Haiku also wins on output tokens.
Side-by-side comparison
| Spec | Anthropic: Claude 3.5 Haiku | xAI: Grok 4 |
|---|---|---|
| Input price (per 1M) | $0.80 | $3.00 |
| Cached input (per 1M) | $0.08 | $0.75 |
| Output price (per 1M) | $4.00 | $15.00 |
| Batch input (per 1M) | $0.40 | — |
| Batch output (per 1M) | $2.00 | — |
| Reasoning price (per 1M) | — | — |
| Context window | 200K | 256K |
| Vision support | Yes | Yes |
| Caching support | Yes | Yes |
| Batch API | Yes | No |
| Reasoning capability | No | No |
Monthly cost at volume
Estimated monthly API spend at common production traffic levels (input/output tokens per request shown).
| Volume | Anthropic: Claude 3.5 Haiku | xAI: Grok 4 | Savings |
|---|---|---|---|
1K req/day 500in / 200out tokens | $36.00 | $135.00 | $99.00 Anthropic: Claude 3.5 Haiku wins |
10K req/day 1500in / 500out tokens | $960.00 | $3,600 | $2,640 Anthropic: Claude 3.5 Haiku wins |
100K req/day 3000in / 800out tokens | $16,800 | $63,000 | $46,200 Anthropic: Claude 3.5 Haiku wins |
1M req/day 8000in / 2000out tokens | $432,000 | $1,620,000 | $1,188,000 Anthropic: Claude 3.5 Haiku wins |
Adjust input/output token counts, request volume, batch & cached pricing.
Related comparisons
Frequently asked questions
Which is cheaper, Anthropic: Claude 3.5 Haiku or xAI: Grok 4?
For input tokens, Anthropic: Claude 3.5 Haiku is roughly 73% cheaper at $0.80/1M vs $3.00/1M. For output tokens, Anthropic: Claude 3.5 Haiku wins at $4.00/1M. Real-world cost depends on your input/output ratio — use the calculator to model your actual workload.
What’s the context window difference?
Anthropic: Claude 3.5 Haiku has a context window of 200K tokens. xAI: Grok 4 offers 256K tokens. Larger context windows are valuable for long documents, RAG pipelines, and multi-turn conversations — but they come with higher input-token bills if you fill them every request.
Should I use Anthropic: Claude 3.5 Haiku or xAI: Grok 4?
Choose Anthropic: Claude 3.5 Haiku if you’re already on the Anthropic stack, want broad ecosystem support, or prefer its lower input price. Choose xAI: Grok 4 for xAI’s ecosystem, native vision input, or its differentiated capabilities. Run a small benchmark on your own prompts before committing — price is only one axis.
How are these prices kept current?
Prices are pulled directly from OpenRouter’s public models API once every 24 hours via a Convex cron job, then normalized to per-1M-token figures. Last refresh: Apr 21, 2026.