VerdictPal
Join

Models / Anthropic

Claude Sonnet 5

Claude Sonnet · Released 2026-06-30

Anthropic's June 2026 Sonnet-class release. Artificial Analysis Intelligence Index 53.4 at max adaptive effort — between GPT-5.5 and Claude Opus 4.7 on the public AA board. API list price $2 / $10 per 1M tokens.

Default Anthropic mid-tier once Sonnet 5 is generally available in your stack. Use max-effort AA rows for intelligence comparisons; lower-effort variants score differently on the same suite.

#13 of 168 on AA Index · current snapshot
55 AA Index · 10 benchmark rows · 2 sourcesPosition is benchmark-specific — not a cross-family or cross-source ranking.
Composite62
Context1M
Input / 1M$2.0
Output / 1M$10
Knowledge cutoffopen
Output speed80 t/s
TTFT (default API)142.07s
Statussolid
Claude

Family profile

Best published score in each covered benchmark family.

76/100 avg
Reasoning
91
Coding
54
Agentic
81
Long context
77

4 tested benchmark families

Editor's note
Tool scaffolded from the 2026-07-09 AA snapshot (slug claude-sonnet-5, max adaptive effort). Desk still needs full editorial depth; do not treat the summary as a hands-on verdict.
Benchmark placements

Where Claude Sonnet 5 places on each public benchmark source that publishes a row. Rank counts every model with a latest row in the same test — not a universal quality score.

BenchmarkFamilyRankScoreSourceDate
AA time to first tokenPerformance#2/ 186Artificial Analysis2026-09-01
Vending-Bench 2Agentic#10/ 10Andon Labs Vending-Bench 22026-08-04
Artificial Analysis Intelligence IndexReasoning#17/ 18755Artificial Analysis2026-09-01
SciCodeCoding#20/ 18054Artificial Analysis2026-09-01
Terminal-BenchAgentic#22/ 18481Artificial Analysis2026-09-01
τ³-Bench BankingAgentic#23/ 10637Artificial Analysis2026-09-01
AA-LCRLong context#27/ 17977Artificial Analysis2026-09-01
Humanity's Last ExamReasoning#27/ 18041Artificial Analysis2026-09-01
GPQA DiamondReasoning#30/ 18591Artificial Analysis2026-09-01
AA output speedPerformance#50/ 187Artificial Analysis2026-09-01
Family context

The three highest-scoring models with pages in each capability family. Where Claude Sonnet 5 shows up, it's highlighted.

ReasoningClaude Sonnet 5 · 91
  1. 1
    Claude Mythos Preview
    95
  2. 2
    Gemini 3.7 Flash
    95
  3. 3
    Grok 4.6
    95
CodingClaude Sonnet 5 · 54
  1. 1
    GPT 5.5
    96
  2. 2
    Claude Opus 4.8
    96
  3. 3
    Claude Opus 4.6
    96
AgenticClaude Sonnet 5 · 81
  1. 1
    GPT 5.5
    100
  2. 2
    Claude Opus 4.8
    100
  3. 3
    Claude Fable 5
    100
Long contextClaude Sonnet 5 · 77
  1. 1
    Muse Spark 1.2
    83
  2. 2
    Kimi K3
    83
  3. 3
    Gemini 3.5 Flash
    81
Newest receipts

Benchmark rows added to the public ledger for Claude Sonnet 5 in the last 120 days. Older rows live in the full table below.

10 recent
BenchmarkFamilyScoreSourceDays ago
AA-LCRLong context77%Artificial Analysis5
AA output speedPerformance80 t/sArtificial Analysis5
AA time to first tokenPerformance142.07sArtificial Analysis5
Artificial Analysis Intelligence IndexReasoning55.3%Artificial Analysis5
GPQA DiamondReasoning91.1%Artificial Analysis5
Humanity's Last ExamReasoning41.3%Artificial Analysis5
SciCodeCoding53.6%Artificial Analysis5
τ³-Bench BankingAgentic37.32%Artificial Analysis5
Terminal-BenchAgentic80.52%Artificial Analysis5
Vending-Bench 2Agentic6377.7Andon Labs Vending-Bench 233
Family coverage

A high composite that hides a weak family is a trap. These bars surface the families where this model hasn't been publicly tested, and where it leads.

Reasoning91
Coding54
Agentic81
Long context77
Price vs. performance
0255075100$0.10$0.30$1$3$10$30input USD / 1M tokens · log scale →← intelligence

Hover any dot for name, score, and input price. Keyboard: tab through the top twelve, or use the ranking below.

AnthropicOpenAIMistralAlibabaDeepSeekGoogleMoonshotMetaMiniMaxxAIZhipuXiaomiOtherMeituanCohereNVIDIA

Full benchmark ledger

Every catalog benchmark for Claude Sonnet 5. Scores link to the original source; gaps mean no public row exists yet.

10 sourced rows

10 of 37 catalog benchmarks have a sourced row for Claude Sonnet 5.

Reasoning

Coding

Agentic

Long context

Performance

Changelog
  1. Claude Sonnet 5 listed on Artificial Analysis (max-effort index 53.4).
Sources