When was Claude Sonnet 5 released?

Claude Sonnet 5 launched on June 30, 2026 with same-day availability across the Claude apps, Claude Code, the Claude API (claude-sonnet-5), AWS Bedrock, Google Vertex, Microsoft Foundry, and third-party tools including Cursor, VS Code, GitHub Copilot, and OpenRouter.

How much does Claude Sonnet 5 cost?

Introductory pricing is $2 per million input tokens and $10 per million output tokens through August 31, 2026, then standard $3 / $15. A new tokenizer counts roughly 30% more tokens for the same text, so real per-task cost can run higher than the per-token rate implies.

Is Claude Sonnet 5 better than Claude Fable 5?

Not on raw capability. Fable 5 is Anthropic’s Mythos-class model and still leads on the longest autonomous work — though since Claude Opus 5 shipped on July 24, 2026 it no longer leads across the board. Sonnet 5 is far cheaper and fast, landing close to Opus 4.8 quality — ideal for high-volume coding and tool use where Fable 5’s $10/$50 price is not justified.

What changes when I migrate from Sonnet 4.6 to Sonnet 5?

Update the model ID to claude-sonnet-5. Adaptive thinking is on by default; manual extended thinking now returns a 400 error; and setting temperature, top_p, or top_k to non-default values returns a 400 error. Priority Tier is not available on Sonnet 5.

Back to blog

Claude Sonnet 5: Release, Pricing & How It Compares to Fable 5

Claude Sonnet 5 (claude-sonnet-5) launched June 30, 2026 as Anthropic's most agentic Sonnet. Pricing, 1M context, migration changes, and when to pick it over Fable 5.

Jul 1, 2026

Fable5 Editorial

On June 30, 2026, Anthropic shipped Claude Sonnet 5 (claude-sonnet-5) — the newest model in its mid-tier Sonnet line and, in the company's framing, its most agentic Sonnet yet. A day later, Claude Fable 5 returned globally from its export-control suspension. Here is what Sonnet 5 is, what it costs, and when to reach for it instead of Fable 5.

Claude Sonnet 5: the short answer

Sonnet 5 plans multi-step work, drives browsers and terminals, and runs autonomously at a level that recently required larger, more expensive models. It ships with a 1M-token context window and up to 128K output tokens, and lands close to Opus 4.8 quality for far less money. See the full spec sheet on the Claude Sonnet 5 model page.

Pricing

Sonnet 5 launches at introductory pricing of $2 / $10 per million input/output tokens through August 31, 2026, then moves to standard $3 / $15 — the same list price as Sonnet 4.6.

One catch: Sonnet 5 uses a new tokenizer that produces roughly 30% more tokens for the same text. So even though the per-token rate is unchanged from 4.6 at standard pricing, an equivalent request can cost more in practice. Budget with the token inflation in mind.

Availability

Day-one availability was unusually broad:

Default model for Free and Pro on claude.ai; available to Max, Team, and Enterprise
Claude Code and the Claude API as claude-sonnet-5
AWS Bedrock, Google Vertex, Microsoft Foundry
Third-party tools: Cursor, VS Code, GitHub Copilot, OpenRouter

Migration: three behavior changes

If you are moving from Sonnet 4.6, update the model ID and note:

Adaptive thinking is on by default.
Manual extended thinking is removed — thinking: {type: "enabled", budget_tokens: N} now returns a 400 error.
Sampling parameters are locked — setting temperature, top_p, or top_k to non-default values returns a 400 error.

Priority Tier is not available on Sonnet 5.

Sonnet 5 vs Fable 5: which to use

Dimension	Claude Sonnet 5	Claude Fable 5
Tier	Mid-tier, agentic	Mythos-class, priced a tier above Opus
Intro price (in / out)	$2 / $10	$10 / $50
Context	1M tokens	1M tokens
Best for	High-volume coding, tool use, browser/terminal agents	Hardest reasoning, multi-day autonomy, expert analysis

Rule of thumb: start on Sonnet 5 for everyday agentic coding — it's five times cheaper on input and output. Escalate to Claude Fable 5 only when peak reasoning quality justifies the cost. Since a well-structured prompt ports across the whole Claude lineup, keep one portable system prompt and switch models per task.

Summary

Released June 30, 2026 as claude-sonnet-5
Introductory $2 / $10 through Aug 31, 2026, then $3 / $15
1M context, 128K output, adaptive thinking on by default
New tokenizer counts ~30% more tokens — budget accordingly
Pick Sonnet 5 for volume and speed; Fable 5 for peak capability

Independent analysis, last updated July 1, 2026. Official sources always win on conflict — see anthropic.com/news for the primary announcements.