Anthropic Releases Claude Sonnet 5 — Near-Flagship Performance at Half the Cost

Anthropic released Claude Sonnet 5 on June 30, 2026, making it the company's most capable mid-tier model to date. The release positions Sonnet 5 as a near-replacement for Opus 4.8 in most workloads — at roughly half the price.

On knowledge work benchmarks, Sonnet 5 actually edges out Opus 4.8. On harder reasoning tasks and deep research, Opus still holds the lead, but the gap has narrowed substantially. Across reasoning, tool use, coding, and general knowledge work, Sonnet 5 is a clear step up from its predecessor, Sonnet 4.6.

Pricing and Availability

Sonnet 5 is available immediately across all Claude plans. It is now the default model for Free and Pro plans, replacing Sonnet 4.6 in that slot.

Through August 31, 2026, Anthropic is offering promotional API pricing: $2 per million input tokens and $10 per million output tokens. After that, standard pricing kicks in at $3 per million input tokens and $15 per million output tokens. For comparison, Opus 4.8 runs at $15 per million input and $75 per million output — making Sonnet 5 at promotional pricing about 87% cheaper on both ends, for a model that, in many practical scenarios, performs comparably.

Built for Agents

Anthropic describes Sonnet 5 as its most agentic Sonnet model yet. It's designed to plan across multi-step tasks, use tools like web browsers and terminals, and run autonomously for extended periods without losing context. This makes it well-suited for AI coding tools, automated research workflows, and any pipeline that chains multiple tool calls together.

The model supports a 1 million token context window — enough to hold an entire large codebase or a lengthy research corpus — and Anthropic's system card reports lower rates of hallucination and sycophancy compared to Sonnet 4.6, along with improved refusal of malicious requests without over-refusing legitimate ones.

Who Should Switch

For developers and organizations currently running Opus 4.8 for anything short of the hardest edge cases, Sonnet 5 is now a serious alternative worth evaluating. At promotional pricing, the cost difference is large enough to justify running Sonnet 5 as the primary model and reserving Opus 4.8 only for tasks where the performance gap is measurably meaningful.

For end users on Free and Pro plans, Sonnet 5 simply becomes the new Claude — no configuration needed. It's already the default.