Claude Sonnet 5.5 Review: Features, Benchmarks & Pricing

Claude Sonnet 5.5 is Anthropic's newest mid-tier model, released on September 28, 2026, just six days after Claude Opus 5.5. It succeeds Claude Sonnet 5 at the same $2/$10 API price, runs more than 30% faster, and in Anthropic's benchmarks lands within a couple of points of Opus 5.5 — and even beats it on one coding test.

This Claude Sonnet 5.5 review covers what changed, what the benchmark numbers show, what it costs, and where its limits sit. It also covers access: Claude Sonnet 5.5 is not yet integrated into Chat Smith, but several other Claude models already are.

Claude Sonnet 5.5 is Anthropic's Sonnet-class model for well-scoped everyday work: coding, agents, and document-heavy tasks. Anthropic describes it as a clear upgrade over Sonnet 5 that shows better judgment across different levels of complexity while spending significantly fewer output tokens.

It supports a 1-million-token context window and is available through the Claude API as claude-sonnet-5-5. In Anthropic's lineup, it sits below Opus 5.5 and above the new Haiku 5.5, with the Fable models above Opus; our Claude Fable 5 overview covers that top tier.

If you have used Sonnet 5, the biggest differences are speed and efficiency: Anthropic says output is generated more than 30% faster, and most tasks cost up to 30% less because the model needs fewer tokens to finish them.

Four areas show the biggest gains over Claude Sonnet 5.

The headline result is Terminal-Bench 4.0, which tests autonomous work in a terminal: Sonnet 5.5 scores 70.6%, up from 10.3% for Sonnet 5 and ahead of Opus 5.5's 66.4%. On CursorBench 4.0 it reaches 55.5%, about 2 points behind Opus 5.5 and far above Sonnet 5's 34.1%. One early customer reported that across 118 real app builds, Sonnet 5.5 produced apps that scored level with Opus 5.

On GDPval-AA v2.1, which measures economically valuable work across professions, Sonnet 5.5 has an Elo of 1,844, effectively tied with Opus 5.5 at 1,846 and well ahead of GPT-6 Sol at 1,487. In one customer's suite of 2,441 finance tasks, it scored higher than Sonnet 5 while using about 121,000 tokens per answer instead of 497,000. That efficiency matters for jobs like planning with an AI business plan generator or analyzing long reports.

Anthropic says Sonnet 5.5 generates output more than 30% faster than Sonnet 5. On several benchmarks, its low or medium effort settings beat Sonnet 5's best score at roughly a tenth of the cost per task, according to Anthropic's score-versus-cost charts. For everyday work, that means you rarely need the highest effort level.

With a 1M-token context window, Sonnet 5.5 can work across large codebases, contracts, or research files in one session. Combined with its lower token use, that makes it a practical choice when you need to summarize a long PDF or compare several documents without paying Opus prices.

The table below uses Anthropic's launch figures. Prices are API rates per million tokens.

The pattern is clear. Sonnet 5.5 closes nearly all of the gap to Opus 5.5 at half the price, and on Terminal-Bench it moves ahead. Sonnet 5 remains a capable model, but its successor is far stronger on agentic coding.

Claude Sonnet 5.5 costs $2 per million input tokens and $10 per million output tokens, unchanged from Sonnet 5 and half of Opus 5.5's $4 and $20. Anthropic offers up to 90% savings with prompt caching and 50% with batch processing. The savings it highlights come from token efficiency rather than a price cut.

• Claude apps: available to anyone on Claude.ai on the web, iOS, and Android.
• Developers: available as claude-sonnet-5-5 on the Claude Platform, Amazon Web Services, Google Cloud, and Microsoft Foundry.
• Chat Smith: Claude Sonnet 5.5 is not integrated yet. You can use Claude Sonnet 5 and other Claude models today, with higher usage limits on Chat Smith Pro at $9.99 per month or $39.99 per year.

Cost per task depends on effort level. Anthropic says most tasks cost up to 30% less than on Sonnet 5, but an independent pre-release evaluation at Max effort estimated about $7.60 per Intelligence Index task, roughly 50% more than Sonnet 5 in that test. Lower effort settings are usually the better value.

It also ships with new safety measures. Sonnet 5.5 is the first Sonnet model with dedicated cybersecurity safeguards: higher-risk security tasks visibly fall back to Claude Sonnet 5, while routine software development is unaffected.

Finally, Opus 5.5 still edges it on CursorBench and some harder professional tasks, and the benchmark figures are Anthropic's own. Test Sonnet 5.5 against alternatives such as GPT-6 Sol on your real tasks before committing.

Claude Sonnet 5.5 is not integrated into Chat Smith yet, but several Claude models already are. Pick one from the model list and start chatting on Android, iOS, or in the browser, with no Anthropic account needed.

Claude Sonnet 5 is the closest option: the same price tier, a 1M-token context window, and strong writing, research, and coding skills. For quick, lightweight tasks, Claude Haiku 4.5 is faster, and Claude Sonnet 4.6 remains available if you have prompts tuned for it. Sonnet 5.5's biggest gains show up mainly in long agentic coding and token-heavy professional work.

Because Chat Smith also includes models from OpenAI, Google, xAI, and DeepSeek, you can put the same prompt to a Claude model and its rivals and keep the answer that holds up. Chat Smith Pro adds higher usage limits and more premium models for $9.99 per month or $39.99 per year.

Claude Sonnet 5.5 is one of the strongest mid-tier models available: near-Opus results on coding and professional work, a big jump on agentic terminal tasks, faster output, and lower cost per task at the same $2/$10 price. Its main caveats are higher costs at maximum effort and new cybersecurity restrictions.

Until it reaches Chat Smith, the Claude models already in the app are an easy way in. You can try them now in Chat Smith, an AI chat app that puts Claude next to the latest models from OpenAI, Google, xAI, and DeepSeek.