Anthropic launches Claude Sonnet 5.5 with 30% faster speed and 30% lower task costs
On September 29, Anthropic released Claude Sonnet 5.5, a mid-tier AI model that runs over 30% faster and reduces task costs by up to 30% compared to Sonnet 5. The model scored 70.6% on Terminal-Bench 4.0, up from 10.3%, and 56 on the Artificial Analysis Intelligence Index, 2 points behind Opus 5.5. Pricing remains at $2 per million input tokens and $10 per million output tokens. The model is available on the Claude platform, AWS, Google Cloud, Microsoft Azure, and OpenRouter.
IllustrationEditorial responsibility
- No named human review is recorded for this page.
- Reports are grouped by semantic similarity and deterministic rules. Language models may assist titles, summaries, translation and cross-source analysis; the page reads the event directly, while its address stays stable when the title changes.
- Summary covers the current reports
Cross-source coverage
Reporting timeline
Claude Sonnet 5.5 roughly matches Claude Fable 5.1, Anthropic mid-size model reaches state of the art
A post on X by user kimmonismus claims that Anthropic's mid-size model, Claude Sonnet 5.5, has roughly matched the performance of Claude Fable 5.1, which scored 165 on an unspecified benchmark. The author states that the mid-size model now matches the state of the art and is probably the best model in the world, achieving this within just a few months of release. The post expresses excitement for a future Fable 5.5 model. This is an attributed opinion and forecast based on the author's assessment, not an official Anthropic announcement.
Anthropic releases Claude Sonnet 5.5 on OpenRouter, claiming 30% faster and cheaper performance
OpenRouter announced that Claude Sonnet 5.5, developed by Anthropic, is now live on its platform. According to the announcement, Sonnet 5.5 is significantly smarter than its predecessor Sonnet 5 and writes much more clearly. The new model is also reported to be 30% faster and approximately 30% cheaper for most workloads compared to the previous version. The post includes a link to the OpenRouter service for access.
Read sourceAnthropic releases Claude Sonnet 5.5 with 70.6% Terminal-Bench score, 30% cost reduction
Anthropic has released Claude Sonnet 5.5, a new AI model that scores 70.6% on Terminal-Bench 4.0, a significant jump from Sonnet 5's 10.3%. The model maintains the same pricing as Sonnet 5 at $2 per million input tokens and $10 per million output tokens, but achieves up to 30% cost reduction per task due to faster speeds and fewer tool calls. Output arrives more than 30% faster than Sonnet 5, making it Anthropic's quickest Sonnet model yet. Users can adjust an effort setting, trading longer reasoning for higher cost per task. Anthropic characterizes Sonnet 5.5 as ideal for routine work such as software debugging, coding, document production, building presentations and spreadsheets, and designing interfaces. The company claims that at Low or Medium effort, Sonnet 5.5 exceeds Sonnet 5's top score at about one-tenth the cost per task, and on FrontierCode at High effort it scores roughly 10 points above Sonnet 5 at approximately one-fifteenth the cost per task.
Read sourceShow 12 older updatesHide older updates
Anthropic launches Claude Sonnet 5.5, scoring 56 on Intelligence Index, 2 points behind Opus 5.5
Anthropic has launched Claude Sonnet 5.5, a new AI model that scores 56 on the Artificial Analysis Intelligence Index, just 2 points behind the top-scoring Opus 5.5. At its max effort setting, Sonnet 5.5 gains 18 points over its predecessor Sonnet 5 and ranks second on the Intelligence Index. The model is priced identically to Sonnet 5 at $0.2/$2/$10 per 1M cache input/input/output tokens, but outputs a higher number of tokens per task, resulting in a cost of $7.60 per task, roughly 50% higher than Sonnet 5. Key benchmarks show Sonnet 5.5 reaching parity with Opus 5.5 on agentic terminal use (Terminal-Bench 4.0: 64% vs 60%) and knowledge work (AA-Briefcase, GDPval-AA, AutomationBench-AA), though with significantly higher token usage. At max effort, it used ~193k output tokens per Intelligence Index task, the highest measured, about 60% higher than Opus 5.5 and 7x GPT-6 Astra. The model lags behind Opus 5.5 on factual knowledge (54% vs 66%) and scientific reasoning (6 points lower on Humanity's Last Exam and SciCode). The evaluations were conducted on a pre-release deployment with a known bug affecting structured outputs, which Anthropic says is fixed for public release. The model retains a 1 million token context window and offers five effort settings.
Read sourceAnthropic Releases Claude Sonnet 5.5, Scores 56 on Artificial Analysis Intelligence Index
Anthropic has released Claude Sonnet 5.5, a new AI model. According to a post on X by Artificial Analysis (@ArtificialAnlys), the model scored 56 points on the Artificial Analysis Intelligence Index. This score is only 2 points lower than the top score of Opus 5.5 (max). In max effort mode, Claude Sonnet 5.5 outperforms its predecessor, Sonnet 5, by 18 points. Artificial Analysis uses its own benchmark to break down the trade-off between performance and token cost, providing a quantifiable reference for model selection.
Read sourceAnthropic's Sonnet 5.5 update fixes Claude Code bug, boosting speed by 30% and cutting usage by 30%
Anthropic released Sonnet 5.5, a bug-fix update for Claude Code, its AI-powered coding assistant. The update delivers a 30% improvement in speed and a 30% reduction in resource usage, according to a post by bcherny on X. The post did not specify the nature of the bug or provide additional technical details. The performance gains are attributed directly to the Sonnet 5.5 update.
Anthropic Launches Claude Sonnet 5.5 with 30% Faster Speed and 30% Lower Cost
On September 29, Anthropic announced the release of its new AI model, Claude Sonnet 5.5. The company claims the model operates over 30% faster than its predecessor, Sonnet 5, and can reduce task execution costs by up to 30%. Positioned as a mid-tier model in the Claude 5.5 series, Sonnet 5.5 is designed for everyday tasks such as code fixes, document creation, presentations, and spreadsheet work. Anthropic reported significant performance improvements across multiple tests, notably scoring 70.6% on the Terminal-Bench 4.0 code test, compared to Sonnet 5's 10.3%. The model also features enhancements in long-cycle tasks, image understanding, and knowledge work. Pricing remains unchanged from Sonnet 5 at $2 per million input tokens and $10 per million output tokens, with cached reads at $0.2 per million tokens, but Anthropic states that efficiency gains can lower actual task costs by up to 30%. Sonnet 5.5 includes upgraded security measures, marking the first time the Sonnet series has adopted cybersecurity protections similar to high-performance models. The model is available on the Claude platform and via Amazon Web Services (AWS), Google Cloud, and Microsoft Azure.
Read sourceAnthropic claims new Sonnet model costs 30% less and is 30% faster than predecessor
An anonymous X user, posting under the handle kimmonismus, highlights a statement from Anthropic regarding its latest Sonnet AI model. According to the post, Anthropic claims the new model costs up to 30% less per task than its predecessor and generates output more than 30% faster, making it the fastest Sonnet model to date. The user notes that the pricing is the same as OpenAI's GPT-6-Sol, but asserts the Anthropic model is much more efficient, faster, and generally better. The user speculates that OpenAI will need to release a very strong model in response, and that Anthropic is strongly competitive again. The user also claims the new Sonnet model is just short of Opus 5.5 but is much faster and cheaper. These are attributed opinions and forecasts, not verified facts.
Read sourceAnthropic Launches Claude Sonnet 5.5 AI Model with 30% Speed Boost and Cost Reduction
On September 29, Anthropic released its new AI model, Claude Sonnet 5.5, claiming it offers over 30% faster operation and up to 30% lower task costs compared to its predecessor, Sonnet 5. Positioned as a mid-range model in the Claude 5.5 series, Sonnet 5.5 is designed for daily tasks such as coding, document creation, presentations, and spreadsheets. Anthropic reported significant performance improvements in various tests, including a score of 70.6% on the Terminal-Bench 4.0 coding benchmark, up from Sonnet 5's 10.3%. The model also features enhancements in long-cycle tasks, image understanding, and knowledge work. Pricing remains unchanged from Sonnet 5 at $2 per million input tokens and $10 per million output tokens, with cached reads at $0.2 per million tokens, but Anthropic states that efficiency gains can reduce actual task costs by up to 30%. The model includes upgraded safety features, marking the first time the Sonnet series has adopted cybersecurity protections similar to high-performance models. Claude Sonnet 5.5 is now available on the Claude platform and supports Amazon Web Services, Google Cloud, and Microsoft Azure.
Read sourceAnthropic announces Claude Sonnet 5.5, 30% faster and cheaper with near-Opus performance
Anthropic has officially announced Claude Sonnet 5.5, a new AI model that is 30% faster and 30% cheaper than its predecessor, while delivering performance close to the more advanced Opus 5.5 model. The announcement positions Sonnet 5.5 as a potential new default option for users, balancing cost efficiency with high capability. The post, shared by the account testingcatalog on X, highlights the model's improved speed and affordability, suggesting it may become a widely adopted choice. No specific release date or pricing details were provided beyond the percentage improvements. The claim of performance near Opus 5.5 indicates a significant upgrade in the Sonnet line, which typically targets a mid-range tier between efficiency and power. This development is part of Anthropic's ongoing competition in the AI model market, aiming to offer more accessible high-performance AI tools.
Read sourceAnthropic launches Claude Sonnet 5.5 with 30% speed boost and 30% cost reduction
Anthropic announced the launch of Claude Sonnet 5.5, a new AI model that is 30% faster and costs up to 30% less than its predecessor, Sonnet 5, for most tasks. The model is described as strong for well-scoped everyday tasks such as fixing bugs, creating documents, and slides. The announcement was made via a post on X by addyosmani, who also noted that @RLanceMartin had the models paint the same photo in code to demonstrate the improvement. The post includes a link to a visual comparison of the models' outputs.
Read sourceAnthropic Releases Claude Sonnet 5.5, Over 30% Faster Than Sonnet 5
Anthropic has released Claude Sonnet 5.5, a new AI model that runs over 30% faster than its predecessor, Sonnet 5. The model features clearer writing and is designed for rapid back-and-forth interactions. It is positioned as distinct from the Claude Opus 5.5 model, which is built for complex judgment tasks. Sonnet 5.5 specializes in well-defined daily tasks such as bug fixing, and creating documents, slides, and tables. The model is available immediately to all users. Additionally, Anthropic announced that Claude Haiku 5.5 will join the series in the coming weeks.
Read sourceClaude Sonnet 5.5: New AI model runs 30% faster, costs 30% less than Sonnet 5
Anthropic has introduced Claude Sonnet 5.5, the second model in the Claude 5.5 family. According to the official announcement from ClaudeAI, the new model is a clear upgrade over its predecessor, Sonnet 5. It runs more than 30% faster and costs up to 30% less for most work. The announcement was made via a post on X, which included a link to further details. This release represents a significant improvement in both performance and cost-efficiency for the Claude model line.
Read sourceClaude Sonnet 5.5 spotted in Claude Code binary, launch expected within hours
A new AI model, Claude Sonnet 5.5, has been identified within the Claude Code binary, according to a post on X by user pankajkumar_dev. The post states that the model has been spotted and that its launch is expected within the next few hours. This indicates an imminent release of an updated version of Anthropic's Claude Sonnet model, likely integrated into the Claude Code development environment. The exact capabilities and improvements of Sonnet 5.5 over previous versions have not been detailed in the post, but the discovery in the binary suggests it is ready for deployment.
Anthropic's Sonnet 5.5 Leaks Ahead of OpenAI DevDay, Claims Show Superior Performance at Lower Cost
According to multiple developer leaks on X, Anthropic's Claude Sonnet 5.5 model has been spotted in gray-scale testing within Claude Code, with a model identifier 'claude-sonnet-5-5' appearing in configuration files. Early benchmarks shared by developers @notjazii and @chetaslua claim Sonnet 5.5 outperforms OpenAI's newly released GPT-6 Sol in coding and UI animation tasks, and approaches GPT-6 Astra in advanced agent capabilities. The leaked pricing is $2 per million input tokens and $10 per million output tokens, significantly undercutting OpenAI's offerings. The article speculates that Anthropic is timing the release to coincide with OpenAI's DevDay on September 29, potentially disrupting the event. Internal data suggests Anthropic's rapid iteration is driven by a data flywheel where 26% of core R&D and 80% of code generation are now AI-automated. The report attributes these claims to unnamed insiders and social media posts, noting the information is unconfirmed by Anthropic or OpenAI.
Read source