Claude Haiku 4.5
Anthropic
37.9#280 of 427 overall
- Input / output
- US$ 1,00 / US$ 5,00 US$ per 1M tokens
- Context
- 200K tokens
- Release
- 10/15/2025 release date
- Weights
- Closed
- Speed
- 0 tok/s first token in 0.00 s
Family history
The 4 versions of this family, from first to newest, by release date. Click a point to open that version.
See the data as a table
| Model | Released | Score | Rank | AA Intel. | Elo Arena |
|---|---|---|---|---|---|
| Claude 3 Haiku | 03/04/2024 | 15.4 | #423 | — | 1195 |
| Claude 3.5 Haiku | 10/22/2024 | 16.4 | #419 | — | 1255 |
| Claude Haiku 4.5 | 10/15/2025 | 37.9 | #280 | — | 1397 |
| Claude Haiku 5.5 | 10/07/2026 | 67.0 | #54 | 36.2 | 1401 |
Scores by source
Scores by sourcepercentile = % of ranked models it beats| Metric | Value | Percentile | Source |
|---|---|---|---|
| Arena text (Elo) weight 2 | 1397 | 44% | LMArena |
| Arena WebDev (Elo) weight 1 | 1330 | 15% | LMArena |
| Humanity's Last Exam weight 1 | 10.4 | 53% | Artificial Analysis |
| Terminal-Bench weight 1 | 0.0 | 22% | Artificial Analysis |
| GPQA Diamond | 67.2 | 46% | Artificial Analysis |
| Long-context reasoning | 74.3 | 72% | Artificial Analysis |
By category
| Category | Score | Rank | Metrics |
|---|---|---|---|
| Agentic | 36.0 | #118 | Terminal-Bench |
| Coding | 32.4 | #101 | Arena WebDev (Elo) |
| Reasoning | 49.6 | #222 | Humanity's Last Exam, GPQA Diamond |
| Research | 61.0 | #93 | Long-context reasoning |
| Design | 32.4 | #99 | Arena WebDev (Elo) |
| Chat | 47.0 | #111 | Arena text (Elo) |
History
RankingIA score on each collection day (hover to see the rank).
Daily collection started on 10/30/2026. The chart appears from the second day the model is ranked.