Claude 4.5 Haiku (Reasoning)
AI model by Anthropic. Real-time pricing and benchmark data.
Benchmarks
Editorial Analysis
Profile. Claude 4.5 Haiku (Reasoning) is positioned in the mid-tier at $5.00/M output tokens. The benchmark profile is balanced — strongest on Coding Index (43.9) — making it a default choice for teams that need general-purpose reasoning without paying frontier-model prices.
Latency. On latency, Claude 4.5 Haiku (Reasoning) reports 89 tok/s output speed and 16690ms time-to-first-token. Output speed is on the slower end; for real-time chat consider pairing with a streaming front-end or routing latency-sensitive requests to a faster model.
Cost model. For a workload of 1M input tokens + 100K output tokens, Claude 4.5 Haiku (Reasoning) costs $1.50 per million input-equivalent requests. Multiply by your monthly request volume to project the bill, and compare against the speed/correctness profile above to decide if the trade-off is worth it for your workload.
Editorial summary generated from public pricing and benchmark data (Artificial Analysis API). All numbers are sourced — see the Artificial Analysis methodology for benchmark definitions. Last refreshed 2026-08-24.
Compare with similar models
| Model | Input | Output | Speed |
|---|---|---|---|
Claude 4.5 Haiku (Reasoning)Current | $1.00 | $5.00 | 89 tok/s |
| Qwen3.5 Omni Plus | $0.40 | $4.80 | 49 tok/s |
| GPT-5.4 mini (medium) | $0.75 | $4.50 | 160 tok/s |
| GPT-5.4 mini (Non-Reasoning) | $0.75 | $4.50 | 162 tok/s |
| GPT-5.4 mini (xhigh) | $0.75 | $4.50 | 163 tok/s |
| o3-mini (high) | $1.10 | $4.40 | 220 tok/s |