Compare Anthropic Claude Sonnet 5, Sonnet 4.6, Opus 4.8, and API pricing.

Just shipped: Anthropic Claude Sonnet 5. The company calls it the “most agentic Sonnet ever”. It runs long tasks, plans and drives terminals.

Sonnet 5 was the model chosen by default for Free and Pro Plans. Max, Team or Enterprise users may select it. The feature is live on Claude Code, as well.

TL;DR

• Anthropic Sonnet 5 is the most agentic model in its mid-tier.The gap between Opus 4.8 and 4.9 is now much smaller.
• Sonnet 4 6 is superior to all benchmarks published: 63.2% SWE-bench Pro, 81.2% OSWorld-Verified, 57.4% HLE.
• Run cheaperPricing for MTok is $2/$10 until Aug 31. After that, it’s $3/$15. Opus 4.8 costs $5/$25.
• The best possible value for low/medium effortsIt is more expensive than Opus 4 for the same quality at xhigh.
• Safety is higher than the 4.6 version, but with a deliberately lower cyber capability — Opus stays the pick for accuracy-critical work.

Claude Sonnet 5

The middle product in Anthropic’s line-up is Sonnet. This is below Opus 4.8 and above Haiku 4.5, the more affordable model.

Sonnet 5 represents an update to Sonnet 4.6 which was released in February 2026. The focus of this release, according to Anthropic, is agentic reliability and not a single headline benchmark.

It means, in reality, longer task chain without losing context. This means that a failed tool call can be corrected more quickly. The behavior will be more stable over the course of extended sessions within Claude Code or Cowork.

Models show effort levels as low, medium high and extra high. A higher effort requires more reasoning tokens. Both quality and costs will increase.

The same tokenizer introduced in Opus 4.7 is used by Sonnet 5. It is possible to map the same text to approximately 1.0 to 1,35 times more symbols.

Explainer

Benchmark

The Anthropic Team published a table of benchmarks comparing Opus 4, Opus 5, and Sonnet 5. Sonnet 5 is superior to its predecessors in all categories. The gap between Opus 4.8 and Sonnet 5 is reduced.

Sonnet 5, on the SWE Bench Pro, scores 63.2%. Sonnet 5 scored 63.2%. Opus 4 8 is the clear leader with 69.2%.

Sonnet 5, based on OSWorld’s verified data, has an 81.2% rating versus Sonnet 46.5. It reaches 80.4% on Terminal-Bench 2.0 compared to 67.0%.

Sonnet 5 scored 57.4% on Humanity’s last exam with tools. This is almost the same as Opus 4.8, which has a score of 57.9%.

On one point, Sonnet 5 is a little bit ahead. Sonnet 5 scores 1 618 points on GDPval AA v2 Knowledge-Work Benchmark, while Opus 4.8 only gets 1,615

The Real Deal: Effort levels

Developers should focus on the story of cost versus performance. Sonnet 5 has a clear improvement on Sonnet 4 at every level of effort. Low and medium efforts show the most value.

Sonnet 5 offers quality at a price that is comparable to earlier Sonnets. Opus 4.8 is the leader in accuracy at the very top.

From this, a routing policy can be derived. Sonnet 5 is the best place to send most of your agentic coding and tool usage, as well as knowledge-based work. Opus 4.8 is reserved for tasks that require accuracy. Haiku is a good choice for those who need to make a lot of calls, but are concerned about latency.

Sonnet 5, Use Cases

The Early Access Partners described specific workflows. The reports are mapped to the common jobs of engineers.

• The Multi-Step Software EngineeringIt was requested by a tester to look into a particular bug. The software wrote a reproducing bug test, applied the fix and then verified that the problem returned. This was done in one pass.
• Brownfield debuggingA second partner used it with hard pull requests. It tracked failures back to the root cause. It sent durable solutions rather than just symptom patching.
• Business automationZapier assigned it a dual-task. The update was made to the Salesforce account levels, and then an email with a link for enterprise contacts was sent. The task was completed from beginning to end.
• Agents that use computersPace automates workflows in insurance, such as the submission of claims and running loss reports. Agents use operational systems already used by teams.
• Data explorationClickHouse agents can query data in real time and generate insights instantly. For analysts, faster reasoning results in faster insight.

Comparative Table

The introductory price of Sonnet 5 is valid until August 31, 2026. After that date, the standard pricing will be $3/15. The standard prompt cache (cache is read at 0.1x input), as well as the Batch API 50% discount, also applies. Sonnet 5 costs less per token than Gemini 3 Pro and GPT-5.5, but more than Gemini 5 Flash. Anthropic’s launch post lists the context window of 1M tokens for Sonnet 5. This post does not include context data for any of the other models.

Coding Example: Calling Sonnet 5

You can call any Anthropic Model. Change the model string. claude-sonnet-5.

Import anthropic

Anthropic = client() # Reads ANTHROPIC_API_KEY

message = client.messages.create(
    model="claude-sonnet-5",
    max_tokens=1024,
    messages=[
        {"role": "user", "content": "Find the race condition in worker.py and ship a tested fix."}
    ],
)

print(message.content[0].text)

Strengths and Weaknesses

Strengths:

• The Anthropic test results show that the Sonnet 4.6 is superior in all benchmark categories.
• Near-Opus 4 quality in several tests, with lower token prices
• On the GDPval AA v2 Knowledge Work Benchmark, Opus 4 8 edges out on its own.
• Sonnet 4:6 has lower hallucination rates, sycophancy and unwelcome-behavior than Sonnet 4.
• Change the API model by simply swapping out the string

Weaknesses:

• The most accurate tasks are still won by Opus 4.
• The cost of Opus 4 at a comparable quality can easily exceed xhigheffort.
• This new tokenizer is capable of increasing token counts up to 1,35 times.
• Opus can be used for approved cyber work. Cyber capacity is deliberately low
• Pricing of 3/15 will become standard after August 31, 2020

Take a look at the Technical details. Also, feel free to follow us on Twitter Don’t forget about our 150k+ML SubReddit Subscribe Now our Newsletter. Wait! What? now you can join us on telegram as well.

Want to promote your GitHub repo, Hugging Face page, Product release or Webinar?? Connect with us