Releases·Model Releasefeatured
Anthropic releases Sonnet 5.5, a cheaper, 30% faster mid-tier model with Opus-level cyber capabilities
Anthropic released Claude Sonnet 5.5, the second model in the Claude 5.5 family, positioning it as a faster, lower-cost complement to Opus 5.5 for well-scoped coding and knowledge work. The company says it runs over 30% faster than Sonnet 5 and effectively costs up to 30% less per task despite identical per-token pricing, and it is the first Sonnet subject to the cybersecurity safeguards applied to Anthropic's most capable models. It is available everywhere, including Amazon Bedrock, and Haiku 5.5 is expected in the coming weeks.
22026 engagement5 sourcesMon, Sep 28, 2026, 06:00 PM
AI Overview
- Anthropic claims Sonnet 5.5 runs 30% faster than Sonnet 5 (released three months earlier) with significantly slower token burn
- Sonnet 5.5 outperforms Opus 5.5 on Anthropic's agentic coding benchmarks, aided by spawning multiple agents within cost limits
- Anthropic says Sonnet 5.5 has cyber capabilities comparable to Opus 5 and is the first Sonnet subject to the same cyber safeguards as Fable and Opus
- Anthropic plans to release a new version of Haiku, its smallest model, in the coming weeks without a firm date
- OpenAI released enhanced versions of Sol and Luna last week, and Meta announced a new model for an upcoming smart glasses feature
- Anthropic positions Sonnet 5.5 as a lower-cost complement to Opus 5.5 for everyday tasks, bug fixes, and polished documents, slides, and spreadsheets
- Sonnet 5.5 keeps the same per-token price as Sonnet 5 but typically needs far fewer tokens for the same work
- An automated behavioral audit shows Sonnet 5.5 improving on Sonnet 5 on most measures of alignment and honesty
Related
Social·Community
Sonnet 5.5 vs Opus 5.5 vs Sonnet 5 in Claude Code and pi: results and behavioral differences on 10 coding tasks
Multiple independent developers benchmarked Anthropic's new Sonnet 5.5 against Opus 5.5 and Sonnet 5 in Claude Code and pi-agent. Sonnet 5.5 matched Opus 5.5's pass rates while running roughly twice as fast and costing one-third to one-half as much, and one tester with skills loaded could not distinguish the two models. A separate comparison found Opus 5.5 itself takes far fewer steps and tokens than Opus 5.
Multiple independent developers benchmarked Anthropic's new Sonnet 5.5 against Opus 5.5 and Sonnet 5 in Claude Code and pi-agent. Sonnet 5.5 matched Opus 5.5's pass rates while running roughly twice as fast and costing one-third to one-half as much, and one tester with skills loaded could not distinguish the two models. A separate comparison found Opus 5.5 itself takes far fewer steps and tokens than Opus 5.
Updated Tue, Sep 29, 2026, 04:23 PM — A skills-focused rerun on 2026-09-29 found Sonnet 5.5 indistinguishable from Opus 5.5 across 3 skills and 3 runs each.
11148 engagement3 sourcesredditThu, Sep 24, 2026, 08:15 PM
Releases·Model Releasefeatured
Anthropic releases Claude Haiku 5.5: 75% cheaper, adjustable effort, first Haiku with effort setting
Anthropic launched Claude Haiku 5.5, its cheapest and most capable small model, priced roughly 75% below Haiku 4.5 on average and 90% less per token on tasks under 100,000 tokens. It targets high-volume, cost-sensitive work like summaries, classification, live support, and browser use, and adds an adjustable effort setting. Anthropic also halved cache-read pricing on Sonnet 5.5.
Anthropic launched Claude Haiku 5.5, its cheapest and most capable small model, priced roughly 75% below Haiku 4.5 on average and 90% less per token on tasks under 100,000 tokens. It targets high-volume, cost-sensitive work like summaries, classification, live support, and browser use, and adds an adjustable effort setting. Anthropic also halved cache-read pricing on Sonnet 5.5.
2.6k engagement2 sourcesreddit, rssWed, Oct 7, 2026, 06:01 PM
Releases·Model Releasefeatured
Sonnet 5.5 launch: pricing matches Sonnet 5 at $2/$10 per million tokens, cache reads at $0.20
A post announces Claude Sonnet 5.5, priced identically to Sonnet 5 at $2/$10 per million tokens. Thinking is always on by default, though developers can use the `between_tools` option to disable upfront thinking. A migration guide is available for teams moving from Sonnet 5.
A post announces Claude Sonnet 5.5, priced identically to Sonnet 5 at $2/$10 per million tokens. Thinking is always on by default, though developers can use the `between_tools` option to disable upfront thinking. A migration guide is available for teams moving from Sonnet 5.
420 engagement1 sourcexMon, Sep 28, 2026, 06:36 PM
Releases·Model Releasefeatured
Anthropic launches Claude Opus 5.5 with Fable-level performance, lower prices, and stricter safeguards
Anthropic released Claude Opus 5.5 on Tuesday, the first model in its Claude 5.5 family and its first release since CEO Dario Amodei called for pacing the frontier. The company says the model matches or beats Fable 5.1 on most benchmarks while costing about 40% less to run, generating output 30%+ faster, and using fewer tokens per task. It ships with Fable-class safety safeguards in biology, cybersecurity, and AI development, with Sonnet 5.5 and Haiku 5.5 due in coming weeks.
Anthropic released Claude Opus 5.5 on Tuesday, the first model in its Claude 5.5 family and its first release since CEO Dario Amodei called for pacing the frontier. The company says the model matches or beats Fable 5.1 on most benchmarks while costing about 40% less to run, generating output 30%+ faster, and using fewer tokens per task. It ships with Fable-class safety safeguards in biology, cybersecurity, and AI development, with Sonnet 5.5 and Haiku 5.5 due in coming weeks.
Updated Wed, Sep 23, 2026, 06:41 AM — OpenAI simultaneously announced GPT-6 Sol and Luna, efficiency-focused models; both vendors pitch 'more for less' — null
1602.9k engagement7 sourcesrss, redditTue, Sep 22, 2026, 04:30 PM
Releases·Tool Release
AWS details running Claude Code on Amazon Bedrock in GovCloud for regulated workloads
An AWS Machine Learning blog post walks through using Claude Code with Claude Opus 5.5 and Sonnet 5.5 on Amazon Bedrock in AWS GovCloud (US), where the models hold FedRAMP Class D certification and support ITAR-regulated development. It gives step-by-step configuration, endpoint selection (bedrock-runtime vs bedrock-mantle), and enterprise deployment guidance for compliance-sensitive teams.
An AWS Machine Learning blog post walks through using Claude Code with Claude Opus 5.5 and Sonnet 5.5 on Amazon Bedrock in AWS GovCloud (US), where the models hold FedRAMP Class D certification and support ITAR-regulated development. It gives step-by-step configuration, endpoint selection (bedrock-runtime vs bedrock-mantle), and enterprise deployment guidance for compliance-sensitive teams.
0 engagement1 sourcerssMon, Oct 5, 2026, 05:25 PM