[good]

Anthropic

Claude Haiku 4.5

Best Value
5.6
out of 10

Anthropic's fastest and most affordable model in the Claude 4 generation, released October 2025. Claude Haiku 4.5 runs at 108.8 tokens/second — fast enough for real-time streaming — at $1/$5 per 1M tokens. Despite the low price, it scores an AA Intelligence Index of 31, placing it #13 of 60 proprietary models. It outperforms Claude Sonnet 4 on computer-use benchmarks (50.7% vs 42.2%) while costing three times less. Supports extended thinking mode (billed at $5/1M for thinking tokens), image input, and the full 200K context window shared across the Claude 4 generation.

Context window

200K tokens

API (blended)

$2.00/1M

Consumer access

Free (limited) / $20/mo

Multimodal

Yes

Try Claude Haiku 4.5Compare

Strengths

  • +Fastest Anthropic model: 108.8 t/s — suitable for real-time streaming and high-throughput pipelines
  • +AA Intelligence Index 31 — top 25% of proprietary models at a budget price point
  • +Outperforms Claude Sonnet 4 on computer-use benchmarks (50.7% vs 42.2%) at 3× lower cost
  • +200K context window — same as Opus and Sonnet 4.6
  • +Extended thinking mode available for complex tasks
  • +Image input supported — describe, analyze, and reason about uploaded images

Weaknesses

  • -Intelligence Index 31 vs Claude Opus 4.6's 53 — substantial capability gap for complex tasks
  • -API output is $5/1M — more expensive per output token than comparable budget models (GPT-5 mini: $2/1M, Gemini 3 Flash: ~$1/1M)
  • -Extended thinking tokens add cost quickly — heavy thinking tasks can approach Sonnet pricing
  • -No native web search; relies on tool use for real-time information

Best for

high-throughput pipelines where speed and cost matterreal-time chatbots and streaming interfacescomputer-use and agentic task automationlightweight coding assistancefast document summarization

Not ideal for

complex multi-step reasoning (use Opus 4.6)nuanced writing and analysis (Sonnet 4.6 clearly better)cost-sensitive output-heavy workloads ($5/1M output is high for budget tier)

Pricing details

Subscription plans

Claude.ai FreeAccess to Haiku 4.5 with daily usage limits(Daily message limits; no API access)
Free
Claude Pro5× more usage than free; access to all Claude models including Opus
$20/mo

API pricing

Anthropic APIStandard pricing: $1/$5 per 1M tokens. Extended thinking mode billed at $5/1M for thinking tokens. 200K context window. Prompt caching available ($0.10/$0.50 per 1M for cache reads). Batch API: 50% discount.
$1/$5

Prices verified February 2026. LLM pricing changes frequently — verify at the provider's site before budgeting.

Last updated: February 2026