SolanaSOL, coin
DevelopingOther

Anthropic Releases Claude Sonnet 5.5, Cheaper and Ahead of Opus 5.5 on Coding

Decrypt reports Anthropic launched Claude Sonnet 5.5 at $2 per million input tokens, scoring 70.6% on Terminal-Bench 4.0 against 66.4% for Opus 5.5.

Why it matters

Sonnet 5.5 undercuts Opus 5.5 on price while scoring higher on coding tests in both Anthropic's figures and the Artificial Analysis figures Decrypt cited, which puts pricing pressure on frontier coding models. The token-use tradeoff is unresolved: Artificial Analysis ranks it second overall and says it consumed more tokens per task than any model it tested. This is a commercial AI model release with no direct effect on crypto asset prices reported in the evidence.

Where the sources disagree

Coding performance of Claude Sonnet 5.5 against Opus 5.5

More tokens per task than any model tested by Artificial Analysis

Decrypt

Nearly a third fewer tokens per task than Opus 5.5

Decrypt

Sonnet 5.5 scores higher on coding tests in both reported sets of numbers, but Artificial Analysis still ranks it second overall behind Opus 5.5, so the lead is limited to the coding test rather than general standing.

Key facts

Anthropic released Claude Sonnet 5.5 on Monday as an upgrade to Sonnet 5.

Disputed figureReported by Decrypt

Anthropic reported a Terminal-Bench 4.0 score of 70.6% for Sonnet 5.5.

70.6 %Disputed figureReported by Decrypt

Anthropic reported a Terminal-Bench 4.0 score of 66.4% for Opus 5.5 and 10.3% for Sonnet 5.

66.4 %Disputed figureReported by Decrypt

Artificial Analysis put Sonnet 5.5 ahead of Opus 5.5, 63.6% to 59.6%.

63.6 %Disputed figureReported by Decrypt

Artificial Analysis ranks Sonnet 5.5 second behind Opus 5.5 overall and reports it used more tokens per task than any model it tested.

Disputed figureReported by Decrypt

Anthropic says Sonnet 5.5 runs more than 30% faster than Sonnet 5 and uses nearly a third fewer tokens per task, making it cheaper to run than Sonnet 5.

30 %Disputed figureReported by Decrypt

What happened

Anthropic released Claude Sonnet 5.5 on Monday, an upgrade to Sonnet 5 from June, Decrypt reported. Pricing held at $2 per million input tokens and $10 per million output tokens, unchanged from Sonnet 5 and half what Opus 5.5 charges. Anthropic said the model runs more than 30% faster than Sonnet 5, and the company wrote that 'Sonnet 5.5 is strongest at well-scoped everyday tasks, fixing bugs, and creating polished documents, slides, and spreadsheets.' On Terminal-Bench 4.0, Anthropic reported 70.6% for Sonnet 5.5 against 66.4% for Opus 5.5 and 10.3% for Sonnet 5. Artificial Analysis, which Decrypt cited as an independent tester, has Sonnet 5.5 ahead 63.6% to 59.6%, but ranks it second behind Opus 5.5 overall and reports it used more tokens per task than any model it tested. Decrypt also noted the model uses nearly a third fewer tokens per task than Opus 5.5, making it cheaper to run than Sonnet 5. The claim comes from a single outlet, and no other testing was reported in the evidence.

How the story developed

  1. First report

    First report by Decrypt

    Anthropic's Claude Sonnet 5.5 Is Out, Beats Opus 5.5 at Coding for Half the Price

    Decrypt

  2. Update

    Anthropic releases Claude Sonnet 5.5

    Launched as an upgrade to Sonnet 5 from June, at unchanged token pricing.

    Decrypt

  3. Official statement

    Anthropic publishes Terminal-Bench 4.0 results

    Anthropic reported 70.6% for Sonnet 5.5, 66.4% for Opus 5.5 and 10.3% for Sonnet 5.

    Decrypt

  4. Update

    Artificial Analysis results reported

    The independent tester put Sonnet 5.5 ahead 63.6% to 59.6% but ranked it second overall behind Opus 5.5.

    Decrypt

  5. Update

    Decrypt reports the release

    Single-outlet account of the launch, pricing and benchmark results.

    Decrypt

Affected entities

Original sources

PublisherReportRolePublished
Decrypt
Crypto media
Anthropic's Claude Sonnet 5.5 Is Out, Beats Opus 5.5 at Coding for Half the PriceOriginal