Claude Haiku 5.5
A small AI language model released by Anthropic, positioned as a cost-efficient, fast-response lightweight model. It supports an adjustable effort parameter and is suited for high-frequency, low-cost large-scale deployment scenarios.
Core Facts
Compared to Claude Haiku 4.5, Haiku 5.5 reduces operating costs by an average of approximately 75%.
Haiku 5.5's input price has dropped to $0.10 per million tokens
Haiku 5.5 is the first Haiku model with an adjustable effort setting.
Timeline (last 90 days)
The v4.1.24 update refreshed the model catalog, adding Claude Haiku 5.5, as well as providers such as AI/ML API, LowRouter, and Prime Inference.
The default models of providers including 302.AI, Cortecs, CrossModel, DevPass, DigitalOcean, Eden AI, GitHub Copilot, Google Vertex AI, LLM Gateway, Meta, Moark, NanoGPT, NEAR AI Cloud, Neuralwatt, Novita AI, Ofox, OpenCode Go, Requesty, and Vivgrid have changed, with most switching to Claude Haiku 5.5 or Muse Wonder 1.0.
Anthropic released Claude Haiku 5.5, with the model ID claude-haiku-5-5, now available on the API, Bedrock, Google Cloud and Copilot.
For all scenarios using Haiku 4.5, Claude Haiku 5.5 is strictly better and 75% cheaper.
Claude Haiku 5.5 has built-in blocking of certain high-risk cybersecurity requests, and therefore is not suitable for scenarios with the highest safety and security requirements.
Claude Haiku 5.5's context window has expanded from Haiku 4.5's 200,000 tokens to 1 million tokens, and the output limit has been raised from 64K to 128K.
Muse Wonder and Claude Haiku 5.5 were released only five days apart.
The key to cost optimization has shifted from 'which model to choose' to 'how to distribute requests', and the reasonable approach is routing: hand about 90% of calls that don't need frontier intelligence to cheap models, and route the remaining 10% of requests that need frontier reasoning to high-end models.
Claude Haiku 5.5 is positioned as a high-throughput scenario model, with applicable scenarios including classification, summarization, information extraction, real-time customer service, voice agents, in-app assistants, and serving as a fast sub-agent model for Opus and Sonnet.
In the per-answer cost including caching, Claude Haiku 5.5 was the lowest at $0.03, Claude Sonnet 5.5 was $0.35, and Gemini 3.8 Flash was the highest at $0.79.
40 more timeline events
All Facts (20)
Compared to Claude Haiku 4.5, Haiku 5.5 reduces operating costs by an average of approximately 75%.
90%VerifiedHaiku 5.5's input price has dropped to $0.10 per million tokens
70%VerifiedHaiku 5.5 is the first Haiku model with an adjustable effort setting.
70%VerifiedClaude Haiku 5.5 is built for high-throughput, cost-sensitive tasks.
65%VerifiedHaiku 5.5 is the first Haiku model to support adjustable effort, allowing developers to select the amount of reasoning resources invested based on task difficulty.
65%VerifiedHaiku 5.5 can work as a sub-agent alongside Claude Opus 5.5 and Sonnet 5.5 to complete programming tasks.
65%VerifiedGemini 2.5 Flash adopts a Mixture of Experts (MoE) architecture, activating only a subset of parameters during inference.
65%Unverified语义缓存、小模型路由、Prompt压缩等技术可以显著降低AI Agent的运营成本
80%UnverifiedThe default models for multiple providers were switched to Claude Haiku 5.5, including Google Vertex AI (changed from Claude Sonnet 5.5 to Claude Haiku 5.5), Cortecs, DevPass, Eden AI, GitHub Copilot, LLM Gateway, NanoGPT, OpenCode Go, Requesty, and Vivgrid.
70%Partially VerifiedAnthropic has released Claude Haiku 5.5, positioned as the lowest-end model in its product line.
65%UnverifiedThe default models of providers including 302.AI, Cortecs, CrossModel, DevPass, DigitalOcean, Eden AI, GitHub Copilot, Google Vertex AI, LLM Gateway, Meta, Moark, NanoGPT, NEAR AI Cloud, Neuralwatt, Novita AI, Ofox, OpenCode Go, Requesty, and Vivgrid have changed, with most switching to Claude Haiku 5.5 or Muse Wonder 1.0.
50%UnverifiedClaude Haiku 5.5's context window has expanded from Haiku 4.5's 200,000 tokens to 1 million tokens, and the output limit has been raised from 64K to 128K.
50%UnverifiedFor all scenarios using Haiku 4.5, Claude Haiku 5.5 is strictly better and 75% cheaper.
50%UnverifiedClaude Haiku 5.5 is positioned as a high-throughput scenario model, with applicable scenarios including classification, summarization, information extraction, real-time customer service, voice agents, in-app assistants, and serving as a fast sub-agent model for Opus and Sonnet.
50%UnverifiedThe key to cost optimization has shifted from 'which model to choose' to 'how to distribute requests', and the reasonable approach is routing: hand about 90% of calls that don't need frontier intelligence to cheap models, and route the remaining 10% of requests that need frontier reasoning to high-end models.
50%UnverifiedClaude Haiku 5.5 has built-in blocking of certain high-risk cybersecurity requests, and therefore is not suitable for scenarios with the highest safety and security requirements.
50%UnverifiedIn the speed test, Claude Sonnet and Haiku were the fastest and most stable, never exceeding 3 minutes; with Mistral and Gemini, about one in every four questions took over 10 minutes, and some approached 25 minutes.
50%UnverifiedClaude Sonnet 5.5's shortcoming is occasionally asking for clarification instead of answering directly; Claude Haiku 5.5's shortcomings are giving partial answers to ambiguous questions and weaker coding ability; GPT-5.6 Terra Pro's shortcomings are asking for clarification when it does not know the asker's identity, and sometimes trusting outdated documentation rather than code.
50%UnverifiedIn the per-answer cost including caching, Claude Haiku 5.5 was the lowest at $0.03, Claude Sonnet 5.5 was $0.35, and Gemini 3.8 Flash was the highest at $0.79.
50%UnverifiedDevin's choice to introduce Haiku 5.5 both as a standalone option and as a Fusion sidekick is a response to the trend of coding agents shifting from performance-only thinking to balancing performance and cost.
50%