Anthropic on Wednesday launched Claude Haiku 5.5, the third model in its Claude 5.5 family, ahead of a planned IPO. The new model targets classification, summarization, and extraction tasks including live support, voice agents, and in-app assistants. It costs 75% less than its predecessor, Haiku 4.5, at $0.10 per million input tokens and $0.50 per million output tokens for prompts under 100,000 tokens, rising to $0.50 and $2.50 for longer prompts.
The pricing announcement came alongside deeper integrations with major cloud providers. Anthropic said Claude Haiku 5.5 is now available through Amazon Web Services, Microsoft Azure, and Google Cloud Platform marketplaces. Each integration includes enterprise-grade security controls that many large organizations require before deploying AI workloads in production.
What's New in the Claude Lineup
Claude Haiku 5.5 is positioned as a cost-effective option for businesses that need reliable AI assistance without the premium price tag of higher-tier models. The model family includes Haiku 4.5, Haiku 5.0, and now Haiku 5.5, each optimized for different use cases. Haiku 5.5 specifically focuses on tasks such as document classification, summarization, and information extraction, with built-in safeguards for high-risk cybersecurity requests.
The company confirmed that Haiku 5.5 is trained on a curated dataset that excludes copyrighted material sourced without permission. This decision directly responds to ongoing legal challenges in the generative AI space, where content creators have sued major developers for training on scraped internet text. Anthropic's legal team, led by General Counsel Andy Bromberg, spent months vetting the training corpus before giving final approval.
According to Deepa Seetharaman in San Francisco, the company emphasized that most everyday tasks would not be impacted by this new release. The pricing structure reflects Anthropic's strategy to make advanced AI more accessible to enterprises while maintaining safety standards. The model is available through Anthropic's API and is also integrated into their Claude Platform, which recently rolled out Fast mode for Claude Opus 5.5 on Google Cloud.
Anthropic CEO Dario Amodei said in a statement that the company's mission has always been to make powerful AI broadly accessible. The 75% price cut, he added, is part of a deliberate effort to remove cost barriers that have kept smaller businesses from experimenting with cutting-edge models. This marks a departure from the industry trend of raising prices as capabilities improve.
Why This Matters Now
The AI market is seeing intense pressure from both cost concerns and regulatory scrutiny. With the EU AI Act enforcement framework imposing fines up to €35 million or 7% of global turnover for non-compliance, companies are under pressure to deploy models that are both efficient and compliant. Anthropic's pricing advantage of 75% over the previous Haiku generation directly addresses enterprise budget constraints while the built-in cybersecurity safeguards help meet regulatory requirements.
The company's compliance features include automated toxicity filtering, bias detection reports, and audit trails that regulators can review. These tools are available at no additional cost to customers on the Pro tier, which starts at $20 per user per month. Enterprises with custom deployments can request additional compliance documentation.
Industry analysts note that the timing is strategic. Anthropic is preparing for an IPO, and this launch is part of a broader effort to expand its AI portfolio before going public. The cost reduction also positions Claude Haiku 5.5 as a competitive alternative to other models from OpenAI and Google, which have been raising prices in recent quarters.
Sarah Chen, a partner at the venture capital firm Redpoint Ventures, told Reuters that the pricing move could accelerate enterprise adoption. She pointed to data showing that 60% of surveyed companies cited cost as the primary barrier to AI deployment. By undercutting competitors by a wide margin, Anthropic may capture market share in sectors like finance, healthcare, and customer service where budget constraints have limited experimentation.
Competitive Context
OpenAI has been focusing on GPT-5 and GPT-5.6, with the latter setting new state-of-the-art results on the Artificial Analysis Coding Agent Index. Google Gemini continues to expand its multimodal capabilities across search and productivity tools. Meanwhile, Anthropic's Haiku 5.5 targets a different segment of the market — cost-sensitive enterprises that need reliable, safe AI without the highest price points.
The Artificial Analysis Coding Agent Index measures model performance on tasks such as code generation, bug fixing, and algorithm implementation. GPT-5.6 Sol with max reasoning scored 80 on this index, setting a new benchmark that other providers are racing to match. Claude Opus 5.5, Anthropic's high-end model, scored 78.5.
Microsoft has been integrating OpenAI models deeply into its Office 365 suite, while Google has bundled Gemini into Workspace and Android. These partnerships have driven adoption but also raised concerns about vendor lock-in. Anthropic's strategy of supporting all three major cloud platforms offers customers more deployment flexibility.
Pricing and Access
The model is priced at $0.10 per million input tokens and $0.50 per million output tokens for prompts under 100,000 tokens, increasing to $0.50 and $2.50 for longer prompts. This tiered pricing structure allows businesses to scale usage based on their needs while keeping costs predictable. The model is available via Anthropic's API and through the Claude Platform, which recently introduced Fast mode for Claude Opus 5.5 on Google Cloud.
For comparison, OpenAI's GPT-4o mini charges $0.15 per million input tokens and $0.60 per million output tokens. Google's Gemini 2.0 Flash costs $0.18 and $0.55 respectively. The pricing differential gives Anthropic a clear advantage in the mid-tier model segment.
Early adopters include Shopify, which is using Claude Haiku 5.5 to power its merchant support chatbot. The company reported a 40% reduction in support ticket resolution time compared to its previous system. Intuit has also integrated the model into TurboTax, where it helps classify tax documents and extract relevant financial information.
Looking Ahead
Anthropic's next steps include expanding the Claude Haiku 5.5 family with additional specialized variants and strengthening enterprise partnerships. The company has also been working on compliance features to help clients navigate the EU AI Act requirements. With the IPO timeline approaching, this launch represents both a product milestone and a strategic move to position Anthropic as a cost-effective leader in the AI space.
Reports suggest that Anthropic's IPO could value the company at $60 billion, making it one of the most highly valued private technology companies in the world. The offering is expected to take place in early 2027, though market conditions may influence the timeline.
Customer Impact and Market Reception
Early feedback from beta testers has been positive, though some users have noted that Haiku 5.5 produces slightly less nuanced output than higher-tier models. The trade-off between cost and quality appears to be acceptable for most enterprise use cases, particularly those focused on classification and data extraction.
Several enterprise software companies have already announced plans to integrate Claude Haiku 5.5 into their products. Salesforce is embedding it into Einstein, ServiceNow reported integration with its Now Assistant platform, and Zendesk is testing it for automated ticket routing.
Financial analysts at Goldman Sachs estimate that the global market for enterprise AI software will reach $120 billion by 2028. Anthropic's aggressive pricing strategy positions it to capture a meaningful share of this market, particularly in sectors where cost efficiency matters more than absolute model quality.
For more on Anthropic's AI developments, see Reuters coverage of the launch