Anthropic Releases Claude Sonnet 5.5, Nearly Matching Opus 5.5 at Lower Cost
Anthropic has launched Claude Sonnet 5.5, the second model in its Claude 5.5 family, delivering more than 30 percent faster output and up to 30 percent lower cost per task while nearly matching the performance of Opus 5.5 on knowledge-work benchmarks. A dramatic jump on the Terminal-Bench coding test signals a major capability upgrade.
Anthropic has released Claude Sonnet 5.5, the second model in its Claude 5.5 family, positioning it as a faster and significantly cheaper alternative that nearly matches the company's flagship Opus 5.5 on knowledge-work benchmarks. The new model generates output more than 30 percent faster and costs up to 30 percent less per task, according to the company.
The most striking improvement comes on Terminal-Bench, a coding benchmark, where Sonnet 5.5 jumps from 10.3 percent to 70.6 percent. That gain suggests a substantial leap in the model's ability to handle complex, multi-step programming tasks, an area where earlier mid-tier models often struggled compared with their larger siblings.
Anthropic's Claude 5.5 family now includes Opus 5.5 and Sonnet 5.5, with Haiku 5.5 announced for the coming weeks. Once Haiku 5.5 arrives, Anthropic will have a direct counterpart to each of OpenAI's three GPT-6 models, setting up a more symmetrical competitive landscape across the major AI labs.
The pricing and speed advantages of Sonnet 5.5 could make it an attractive option for businesses and developers who need strong performance without the higher costs associated with flagship models. By narrowing the performance gap between its mid-tier and top-tier offerings, Anthropic is betting that cost efficiency and throughput will matter as much as raw capability for many real-world deployments.
The benchmark results also highlight how quickly capabilities are advancing within a single model generation. A jump of more than 60 percentage points on Terminal-Bench between versions is unusual and may reflect architectural improvements or training refinements that Anthropic has not fully detailed. The company has not yet published a full technical breakdown of what changed under the hood.
For the broader AI market, the release intensifies competition on price and performance rather than just raw scale. With three tiers of models in the pipeline, Anthropic is clearly aiming to offer customers a clear upgrade path from lightweight to frontier-level systems, all within the same family and presumably with consistent tooling and safety features.
Developers and enterprises evaluating AI providers will now weigh Sonnet 5.5's near-flagship performance against its lower cost per task, potentially shifting more workloads to the mid-tier model. That dynamic could pressure rivals to adjust their own pricing or accelerate their release cycles.
Anthropic has not disclosed exact pricing figures or a precise release date for Haiku 5.5 beyond the coming weeks. The company also has not said whether Sonnet 5.5 will be available through all its existing channels immediately or rolled out gradually.
3
