Claude Haiku 4.5
Anthropic · Released 15 Oct 2025 · claude-haiku-4-5-20251001
$1.00
Input price / 1M tokens
#24 of 61, cheapest first
$5.00
Output price / 1M tokens
#31 of 61, cheapest first
200K
Context window
#62 of 71
91
Output tokens / second
#33 of 61
73.3%
SWE-bench Verified
#7 of 12
Summary
Claude Haiku 4.5 is a proprietary model from Anthropic, released on 15 Oct 2025.
At $1.00 input and $5.00 output per million tokens, it is #24 of 61 on input price, cheapest first.
Its 200K-token context window ranks #62 of 71.
Measured output speed is 91 tokens per second, #33 of 61.
Anthropic's fastest model with near-frontier intelligence; Anthropic positions it as giving coding performance similar to Claude Sonnet 4 at one-third the cost and more than twice the speed.
Benchmarks · 2 reported
| Benchmark | Score | Bar | Rank | Source |
|---|---|---|---|---|
| SWE-bench Verifiedaveraged over 50 trials, no test-time compute, 128K thinking budget | 73.3% | #7 of 12 | Anthropic Haiku 4.5 launch postSelf-reported | |
| Terminal-BenchTerminus 2 agent, XML parser, 32K thinking budget, average of 5 runs (40.21% over 6 runs without thinking) | 41.75% | – | Anthropic Haiku 4.5 launch post (footnote)Self-reported |
Self-reported means the lab ran the test itself. A rank appears only where several labs report the same version of a benchmark.
Similar models
Models with the most benchmarks in common with Claude Haiku 4.5, and the closest scores.
SWE-bench Verified
#7 of 12Real GitHub issues, fixed and tested
- Claude Haiku 4.573.3
- GPT-574.9
- Claude Sonnet 4.577.2
- o369.1
- Mistral Medium 3.577.6
- Gemini 3 Flash78
Compare Claude Haiku 4.5 side by side
Pre-filled with the two closest models. Swap any of them.
Sources
- price, cache prices, batch: platform.claude.com/docs/en/about-claude/pricing, checked 2 Oct 2026
- context, max output, release, api id and alias, knowledge cutoff, modalities: platform.claude.com/docs/en/models/haiku-4-5/overview, checked 2 Oct 2026
- status, retirement commitment (not sooner than 2026-10-15): platform.claude.com/docs/en/about-claude/model-deprecations, checked 2 Oct 2026
- minimum cacheable prompt length: platform.claude.com/docs/en/build-with-claude/prompt-caching, checked 2 Oct 2026
- benchmarks, positioning: anthropic.com/news/claude-haiku-4-5, checked 2 Oct 2026
- output speed: artificialanalysis.ai/models/claude-4-5-haiku, checked 2 Oct 2026
Building on Claude Haiku 4.5?
Get the architecture right before the bill arrives.
We size caching, routing and fallbacks for your workload on a free call, and tell you where a cheaper model is good enough.