Claude Haiku 4.5 Reasoning: pricing, context window and benchmarks
Claude Haiku 4.5 Reasoning is a multimodal model from Anthropic. It offers a 128,000-token context window, costs $15.11 per million input tokens and $70.05 per million output tokens, and averages 81.4 across the index's benchmark suite.
Pricing
Claude Haiku 4.5 Reasoning costs $15.11 per million input tokens and $70.05 per million output tokens. A blended 3:1 input-to-output mix works out to about $28.84 per million tokens. Cached input is listed at $2.19 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $15.11 | per 1M tokens |
| Output | $70.05 | per 1M tokens |
| Cached input | $2.19 | per 1M tokens |
| Blended 3:1 | $28.84 | estimate per 1M tokens |
| 1M in + 1M out | $85.16 | estimate |
Benchmarks
Scores below are modelled estimates on a 0–100 scale. Use them to compare relative position, not as official vendor results.
| Benchmark | Score | Rank | Percentile | Relative |
|---|---|---|---|---|
| MMLU-Pro | 83.4 | 96 | 84th | |
| GPQA Diamond | 74.6 | 150 | 75th | |
| SWE-bench Verified | 69.6 | 119 | 80th | |
| HumanEval | 88.7 | 61 | 90th | |
| MATH-500 | 85.8 | 95 | 84th | |
| AIME 2025 | 79.0 | 100 | 83th | |
| LiveCodeBench | 73.8 | 111 | 82th | |
| IFEval | 89.9 | 73 | 88th | |
| TruthfulQA | 82.5 | 116 | 81th | |
| HellaSwag | 90.9 | 64 | 89th | |
| MMMU | 76.7 | 117 | 81th | |
| Arena Elo | 1444.2 | 90 | 85th |
What Claude Haiku 4.5 Reasoning is good for
Claude Haiku 4.5 Reasoning is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $15.11 per million input tokens, it sits in the mid-range of the index, and the 128,000-token context window is enough for large files and multi-file diffs.
Strengths
- Large 128,000-token context window for long documents and agent traces.
- Accepts more than text, including image input where the provider supports it.
- Reasoning-oriented profile, useful for multi-step problems and structured analysis.
- Coding benchmarks are a relative strength within its tier.
Trade-offs
- Proprietary licence: you cannot download the weights or self-host the model.
- Premium input pricing at $15.11 per million tokens.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Claude 3.7 Sonnet | Anthropic | 200000 | $14.91 | 81.6 |
| Doubao 1.5 Pro Reasoning | ByteDance | 128000 | $14.56 | 81.6 |
| Gemini 2.5 Pro | Google DeepMind | 200000 | $15.20 | 81.6 |
| Llama 4 Maverick | Meta AI | 200000 | $5.36 | 81.6 |
| Claude Haiku 4.5 Reasoning | Anthropic | 128000 | $15.11 | 81.4 |
| Command A Pro | Cohere | 200000 | $14.28 | 81.3 |
FAQ
How much does Claude Haiku 4.5 Reasoning cost?
Claude Haiku 4.5 Reasoning is listed at $15.11 per million input tokens and $70.05 per million output tokens. Cached input is $2.19 per million tokens. A blended 3:1 input-to-output mix is about $28.84 per million tokens.
What is the context window of Claude Haiku 4.5 Reasoning?
Claude Haiku 4.5 Reasoning has a context window of 128,000 tokens and can emit up to 128,000 output tokens per request in this index.
Is Claude Haiku 4.5 Reasoning open weights?
No. Claude Haiku 4.5 Reasoning is a proprietary model with a proprietary licence, so you cannot download the weights.
How does Claude Haiku 4.5 Reasoning score on benchmarks?
Claude Haiku 4.5 Reasoning averages 81.4 across the scored benchmarks in this index, ranking #95 of 600 models. Its strongest relative result is Arena Elo at 1444.2.
What is Claude Haiku 4.5 Reasoning best for?
Claude Haiku 4.5 Reasoning is tuned toward coding work. It is a reasonable candidate for code generation, review, debugging and repository-scale tasks, especially when paired with a retrieval layer. At $15.11 per million input tokens, it sits in the mid-range of the index, and the 128,000-token context window is enough for large files and multi-file diffs.