Sulat.com
AI models
$10 off the fastest DeepSeek V4.1 Flash, Kimi K3 and GLM 5.3 from Synthetic
Vertex logo

Model details

Claude Haiku 4.5

Claude Haiku 4.5 is Anthropic's fastest small model, designed to deliver near-frontier performance at a fraction of the cost. It matches Claude Sonnet 4's capabilities on coding, computer use, and agentic tasks, and actually surpasses it on certain functions like computer interaction. Scoring 73.3% on SWE-bench Verified places it among the world's top coding models. Built for speed and efficiency, Haiku 4.5 excels in real-time applications—chat assistants, customer service agents, pair programming, multi-agent workflows, and rapid prototyping all benefit from its responsiveness, making high-intelligence AI practical for latency-sensitive use cases that previously required larger, costlier models.

The Haiku family traces a lineage of steady improvement, with Haiku 3.5 already surpassing Opus 3 on many intelligence benchmarks just a year earlier. Haiku 4.5 continues that trajectory, bringing that same leap in capability to production environments. The model powers applications like Claude for Chrome, integrates into developer tools including Claude Code and Cursor, and is accessible across multiple cloud platforms including Amazon Bedrock and Microsoft Foundry. With pricing at roughly one-third of Sonnet 4's cost and over twice the speed, it offers developers a practical path to deploy intelligent AI in cost-sensitive, high-throughput production scenarios without sacrificing the coding and reasoning quality that power modern software development.

Vertexclaude-haiku-4-5@20251001claude-haiku

Quick Info

Powered by
Provider
Vertex
Model key
claude-haiku-4-5@20251001
Release date
Oct 15, 2025
Last updated
Oct 15, 2025
Knowledge cutoff
2025-02-28
AI SDK package
@ai-sdk/google-vertex/anthropic
Input modalities
Output modalities
Capabilities

Cost

Input token cost
$1.00
Output token cost
$5.00

Limits

Output tokens
64,000 tokens
Context window
200,000 tokens

Transparent token rates

Compare Claude Haiku 4.5 pricing

Rates are shown per one million tokens. Combined means one million input plus one million output tokens.

Browse this family

Latest news about Claude Haiku 4.5

Vertex

CoverageBenchmark

Claude Haiku 4.5 is positioned as Anthropic's "fastest and most affordable compact model," delivering SWE-bench Verified coding performance of 73.3% — nearly on par with the previous Sonnet 4 (72.7%) — at one-third the cost and roughly twice the speed. Response times feel instantaneous: more than 2x faster than Sonnet The model scores 50.7% on OSWorld for computer use and agent performance. Pricing is $1 input / $5 output per million tokens (USD, excl. tax). Trade-offs include a 200K token context window (vs. 1M in Opus/Sonnet) and a July 2025 training data cutoff. The recommendation: use Haiku for high-speed, high-volume workloads

Vertex

CoverageComparison

Claude Haiku 4.5 is faster and cheaper; Sonnet 4.5 offers deeper reasoning and higher accuracy. This comparison breaks down benchmarks, pricing, and which model fits your use case.

Vertex

CoverageBenchmark

Anthropic released Claude Haiku 4.5, making the model available to all users as its latest entry in the small, fast model category. The company positions the new model as delivering performance levels

Vertex

CoverageBenchmark

Cotera published a real-world agent benchmark of Claude Haiku 4.5 across five production-style tasks, pitching the model as Anthropic's smallest, fastest, cheapest Claude aimed at high-volume agent loops where Sonnet's bill is too high. Haiku 4.5 passed 4 of 5 benchmarks with $0.941 spent in total across 17 tool calls, The post also documents a concrete failure mode: on a Crunchbase research task for Hightouch, Haiku 4.5 emitted a single-sentence preamble ("I'll help you research Hightouch...") and stopped before making any tool calls, producing nothing at a cost of $0.007—described as a known small-Claude pattern where polite-helper

Videos about Claude Haiku 4.5

More models around Claude Haiku 4.5