Observed Signal · Sep 28, 2026 · Product Launch · Source: CNBC Technology · Impact: 4/5 · Sentiment: Positive
Anthropic launches cheaper AI model Sonnet 5.5
Anthropic expanded its Claude 5.5 series with Sonnet 5.5 and Haiku 5.5, featuring cost reductions and performance gains. Sonnet 5.5 is 30% faster and up to 30% cheaper, priced at $2/$10 per million tokens. Haiku 5.5, designed for high-volume agentic tasks, is priced at $0.10/$0.50 per million tokens (under 100K tokens), scores 43 on the Artificial Analysis Intelligence Index (surpassing GPT-6 Luna and Gemini 3.8 Flash), supports 1M-token context, and introduces effort settings. A new tokenizer counts 30% more tokens, but costs are 75% lower. Both models are available on major platforms with integrations. Other AI news includes OpenAI's release of 719 math papers from an unreleased model, Mistral's Le Chonk and Reflection AI's Beam open-weight models, an OpenAI researcher's resignation, Google's public SynthID, OpenAI's EU watermarking, and industry moves like TikTok's shopping assistant and ElevenLabs' valuation doubling.
This is a major AI model release from Anthropic, indicating a strategic move to offer cost-efficient models while addressing safety concerns. It reflects broader industry trends in AI pricing and safety alignment, with implications for AI adoption in advertising/marketing technologies.
Track Anthropic Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your Polaris7 workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Anthropic released Claude Sonnet 5.5 ($2/$10 per million tokens) and Haiku 5.5 ($0.10/$0.50 per million tokens for inputs under 100K tokens), with Haiku up to 90% cheaper for such requests.
- Haiku 5.5 scores 43 on the Artificial Analysis Intelligence Index, surpassing GPT-6 Luna (38) and Gemini 3.8 Flash (41), and supports a 1M-token context with effort settings; a new tokenizer counts 30% more tokens, but costs are 75% lower.
- Both models are available on Claude, AWS, Google Cloud, and Azure; Anthropic cut Sonnet cache-read prices and added integrations with Cursor, GitHub Copilot, and Devin.
- OpenAI published 719 math manuscripts from an unreleased model, and an OpenAI safety researcher resigned; Google opened SynthID publicly and OpenAI will watermark ChatGPT text in the EU.
- Mistral released Le Chonk (1T params) and Reflection AI unveiled Beam (501B MoE); TikTok rolled out an AI shopping assistant, ElevenLabs' valuation doubled to $22B, and China's Manus raised over $500M.
Connected Companies & Entities
“Anthropic on Monday released a new AI model called Sonnet 5.5.”
“Cloud distribution partners include Microsoft, Amazon, and Google.”
“The model is available on all platforms on Monday, including across Amazon Web Services.”
“Google made SynthID publicly available for watermarking AI-generated content.”
“The model is available on all platforms on Monday, including across Google Cloud.”
“The model is available on all platforms on Monday, including across Microsoft Azure.”
“Availability through Claude API as well as Amazon Bedrock, Google Cloud and Microsoft Foundry.”
“OpenAI published math manuscripts and safety researchers resigned.”
“Cursor claims 10x less than Haiku 4.5 on shorter requests and published CursorBench comparisons.”
“Cursor claims 10x less than Haiku 4.5 on shorter requests and published CursorBench comparisons.”
“GitHub Copilot in VS Code reports it matched Claude Sonnet 5 on many coding tasks while using fewer tokens and steps.”
“Brooklyn-based Reflection AI unveiled the Beam open-weight model.”
“TikTok rolls out an AI shopping assistant and one-click checkout capabilities.”
“ElevenLabs valuation doubles to $22 billion amid strong voice AI demand.”
“Semianalysis acquired Citrini Research to expand its market intelligence offerings.”
“Meta products continue to evolve within the competitive social and messaging ecosystem.”
“Participation from other backers including Nvidia.”
“Participation from other backers including Union Square Ventures.”
“Participation from other backers including Menlo Ventures.”
“Participation from other backers including Samsung.”
“French company Mistral released Mistral Large 4, nicknamed Le Chonk.”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Anthropic launches Claude Sonnet 5, cheaper agentic LLM
Anthropic announced Claude Sonnet 5, a midsize, more agentic foundation model intended to run autonomous agent workflows at lower cost. Claude Sonnet 5 can plan, use tools (browsers, terminals) and run end-to-end tasks; it will be the default model for Anthropic’s free and Pro plans. Introductory pricing runs at $2 per million input tokens and $10 per million output tokens through August 31, rising to $3 per million input afterward. Anthropic says Sonnet 5 approaches Opus 4.8’s performance on many tasks while costing less; on an agentic coding benchmark Sonnet 5 scored 63.2% versus Opus 4.8’s 69.2% and Sonnet 4.6’s 58.1%. The company reports safety improvements over Sonnet 4.6, though Opus models and Claude Mythos Preview are still stronger on alignment and dangerous-task resistance. Testers including Zapier highlighted Sonnet 5’s ability to finish complex automation end-to-end.
Anthropic's Sonnet 5.5 Overtakes OpenAI's GPT-6 in AI Ranking
Anthropic has released Claude Sonnet 5.5, a midrange AI model that scores 56 points on the Artificial Analysis Intelligence Index, ranking second overall behind its own flagship Opus 5.5 (58 points) and ahead of OpenAI's GPT-6 Astra (53) and GPT-6 Sol (48). Sonnet 5.5 shows an 18-point improvement over its predecessor Sonnet 5, excelling in agentic tasks and office work, nearly matching Opus 5.5, though it lags in factual knowledge. However, the model consumes significantly more tokens per task (about 193,000 in its highest reasoning mode), making it more expensive per task despite the same list price of $2 per million input tokens and $10 per million output tokens. Anthropic claims up to 30% cost reduction for most work due to efficiency. The model is available on major cloud platforms, and Anthropic has introduced distillation safeguards for the first time on a Sonnet model.
Anthropic Unveils Sonnet 4.6: A Game-Changer in AI
Anthropic announced Sonnet 4.6, a midsized foundation model update released on its four-month cadence. Sonnet 4.6 will become the default model for Free and Pro plan users. The beta offers a 1 million token context window — double Sonnet’s prior maximum — which Anthropic says can contain entire codebases, long contracts or many research papers in a single request. The company highlighted improvements in coding, instruction-following and computer use. Sonnet 4.6 set new record benchmark scores on tasks such as OS World and SWE-Bench and achieved 60.4% on ARC-AGI-2, outperforming most comparable models though trailing some top-tier models like Opus 4.6, Gemini 3 Deep Think and a refined GPT 5.2 variant. Opus 4.6 launched two weeks earlier and an updated Haiku model is expected soon.
Receive all future signals for Anthropic & competitors in your briefing
Never miss material M&A deals, leadership changes, account wins, or product rollouts. Polaris7 monitors 20,000+ primary sources 24/7, delivering verified telemetry directly into your executive briefing.