notable
Anthropic drops its economy recommendation for claude-haiku-4-5 as claude-haiku-5-5 is listed
Anthropic's models page no longer carries the sentence "The fastest model with near-frontier intelligence" that positioned claude-haiku-4-5 as its economy model, and the claude-haiku-4-5 row (input 1, output 5 per MTok) left the compare table in the same fetch. On the same page claude-haiku-5-5 was added at input 0.1 and output 0.5 per MTok with a 1M tokens context window, described "For high-volume, latency-sensitive tasks such as classification, extraction, and routing". The vendor is re-pointing developers in the cost-sensitive tier.
platform.claude.comd8d63d927a62fc72c42eb8377e8832335e3c0384d464f863bce0a1a3daf5f844
Cited change rows
Every field of every cited row, before and after.
| field | before | after |
|---|---|---|
| basis | The fastest model with near-frontier intelligence | The fastest model with near-frontier intelligence |
| class | economy | economy |
| model_slug | claude-haiku-4-5 | claude-haiku-4-5 |
| present | true | false |
| provider | anthropic | anthropic |
| field | before | after |
|---|---|---|
| cached_input_per_mtok | ∅ | ∅ |
| context_window | ∅ | 1M tokens |
| currency | ∅ | USD |
| effective_from | ∅ | ∅ |
| effective_until | ∅ | ∅ |
| input_per_mtok | ∅ | 0.1 |
| model_slug | ∅ | claude-haiku-5-5 |
| notes | ∅ | For high-volume, latency-sensitive tasks such as classification, extraction, and routing. Printed as a "From" price: the lowest band, higher for longer prompts |
| output_per_mtok | ∅ | 0.5 |
| provider | ∅ | anthropic |
| tier | ∅ | ∅ |
| field | before | after |
|---|---|---|
| cached_input_per_mtok | ∅ | ∅ |
| context_window | 200K tokens | ∅ |
| currency | USD | ∅ |
| effective_from | ∅ | ∅ |
| effective_until | ∅ | ∅ |
| input_per_mtok | 1 | ∅ |
| model_slug | claude-haiku-4-5 | ∅ |
| notes | The fastest model with near-frontier intelligence | ∅ |
| output_per_mtok | 5 | ∅ |
| provider | anthropic | ∅ |
| tier | ∅ | ∅ |