Claude Sonnet 4

anthropic/claude-4-sonnet-20250522

Claude Sonnet 4 is an Anthropic model that accepts image, text, and file inputs and generates text output. It provides a context window of 1 000 000 tokens, costs $3 per M input tokens and $15 per M output tokens, and includes support for tools and reasoning. The model is closed‑source and was released on 2025‑05‑14.

  • tool use
  • reasoning

At a glance

provideranthropic
releasedMay 14, 2025
context1M
input / M$3

Specifications

What the API gives you

Context window
1M tokens
Max output
64K tokens
Input price
$3 / 1M tokens
Output price
$15 / 1M tokens
Cache read
$0.30 / 1M tokens
Modalities
image, text, file → text
Knowledge cutoff
Jan 31, 2025
Released
May 14, 2025

Measured evidence

Public benchmark record

Aider Polyglot
61.3%2026-07-17
Aider Polyglot
61.3%2026-07-21
Aider Polyglot
61.3%2026-07-23
Aider Polyglot
61.3%2026-07-27
Aider Polyglot
61.3%2026-07-28
Aider Polyglot
61.3%2026-07-29
Aider Polyglot
61.3%2026-07-30
Aider Polyglot
61.3%2026-07-31
Aider Polyglot
61.3%2026-08-10
Aider Polyglot
61.3%2026-08-01
Aider Polyglot
61.3%2026-08-02
Aider Polyglot
61.3%2026-08-05
Aider Polyglot
61.3%2026-08-03
Aider Polyglot
61.3%2026-08-09
Aider Polyglot
61.3%2026-08-04
Aider Polyglot
61.3%2026-08-06
Aider Polyglot
61.3%2026-08-07
Aider Polyglot
61.3%2026-08-14
Aider Polyglot
61.3%2026-08-08
Aider Polyglot
61.3%2026-08-11
Aider Polyglot
61.3%2026-08-18
Aider Polyglot
61.3%2026-08-12
Aider Polyglot
61.3%2026-08-19
Aider Polyglot
61.3%2026-08-13
Aider Polyglot
61.3%2026-08-15
Aider Polyglot
61.3%2026-08-20
Aider Polyglot
61.3%2026-08-16
Aider Polyglot
61.3%2026-08-17
Aider Polyglot
61.3%2026-08-21
Aider Polyglot
61.3%2026-08-27
Aider Polyglot
61.3%2026-08-22
Aider Polyglot
61.3%2026-08-23
Aider Polyglot
61.3%2026-08-24
Aider Polyglot
61.3%2026-08-25
Aider Polyglot
61.3%2026-09-05
Aider Polyglot
61.3%2026-08-26
Aider Polyglot
61.3%2026-08-28
Aider Polyglot
61.3%2026-09-06
Aider Polyglot
61.3%2026-08-29
Aider Polyglot
61.3%2026-08-30
Aider Polyglot
61.3%2026-08-31
Aider Polyglot
61.3%2026-09-01
Aider Polyglot
61.3%2026-09-02
Aider Polyglot
61.3%2026-09-03
Aider Polyglot
61.3%2026-09-04
Design Arena Coding
1167 Elo2026-07-28
Design Arena Coding
1167 Elo2026-07-27
Design Arena Coding
1167 Elo2026-07-30
Design Arena Coding
1167 Elo2026-07-29
Design Arena Coding
1167 Elo2026-07-31
Design Arena Coding
1166 Elo2026-08-02
Design Arena Coding
1165 Elo2026-08-05
Design Arena Coding
1166 Elo2026-08-01
Design Arena Coding
1164 Elo2026-08-09
Design Arena Coding
1166 Elo2026-08-03
Design Arena Coding
1166 Elo2026-08-04
Design Arena Coding
1161 Elo2026-08-14
Design Arena Coding
1165 Elo2026-08-06
Design Arena Coding
1164 Elo2026-08-10
Design Arena Coding
1165 Elo2026-08-07
Design Arena Coding
1155 Elo2026-08-18
Design Arena Coding
1164 Elo2026-08-08
Design Arena Coding
1155 Elo2026-08-15
Design Arena Coding
1164 Elo2026-08-11
Design Arena Coding
1163 Elo2026-08-12
Design Arena Coding
1161 Elo2026-08-13
Design Arena Coding
1154 Elo2026-08-19
Design Arena Coding
1161 Elo2026-08-27
Design Arena Coding
1154 Elo2026-08-20
Design Arena Coding
1155 Elo2026-08-16
Design Arena Coding
1153 Elo2026-09-05
Design Arena Coding
1155 Elo2026-08-17
Design Arena Coding
1161 Elo2026-08-21
Design Arena Coding
1161 Elo2026-08-22
Design Arena Coding
1161 Elo2026-08-23
Design Arena Coding
1160 Elo2026-08-28
Design Arena Coding
1161 Elo2026-08-24
Design Arena Coding
1161 Elo2026-08-25
Design Arena Coding
1161 Elo2026-08-26
Design Arena Coding
1153 Elo2026-09-06
Design Arena Coding
1160 Elo2026-08-29
Design Arena Coding
1160 Elo2026-08-30
Design Arena Coding
1160 Elo2026-08-31
Design Arena Coding
1160 Elo2026-09-01
Design Arena Coding
1159 Elo2026-09-02
Design Arena Coding
1157 Elo2026-09-03
Design Arena Coding
1156 Elo2026-09-04
Design Arena UI
1162 Elo2026-07-28
Design Arena UI
1163 Elo2026-07-27
Design Arena UI
1162 Elo2026-07-30
Design Arena UI
1163 Elo2026-07-29
Design Arena UI
1162 Elo2026-07-31
Design Arena UI
1163 Elo2026-08-02
Design Arena UI
1162 Elo2026-08-01
Design Arena UI
1161 Elo2026-08-05
Design Arena UI
1158 Elo2026-08-09
Design Arena UI
1162 Elo2026-08-03
Design Arena UI
1161 Elo2026-08-04
Design Arena UI
1156 Elo2026-08-14
Design Arena UI
1160 Elo2026-08-06
Design Arena UI
1160 Elo2026-08-07
Design Arena UI
1158 Elo2026-08-10
Design Arena UI
1158 Elo2026-08-08
Design Arena UI
1153 Elo2026-08-18
Design Arena UI
1158 Elo2026-08-11
Design Arena UI
1154 Elo2026-08-15
Design Arena UI
1159 Elo2026-08-12
Design Arena UI
1159 Elo2026-08-13
Design Arena UI
1153 Elo2026-08-19
Design Arena UI
1153 Elo2026-08-16
Design Arena UI
1153 Elo2026-08-20
Design Arena UI
1151 Elo2026-08-27
Design Arena UI
1153 Elo2026-08-17
Design Arena UI
1147 Elo2026-09-05
Design Arena UI
1152 Elo2026-08-21
Design Arena UI
1152 Elo2026-08-22
Design Arena UI
1152 Elo2026-08-23
Design Arena UI
1152 Elo2026-08-24
Design Arena UI
1151 Elo2026-08-28
Design Arena UI
1152 Elo2026-08-25
Design Arena UI
1153 Elo2026-08-26
Design Arena UI
1148 Elo2026-09-06
Design Arena UI
1151 Elo2026-08-29
Design Arena UI
1152 Elo2026-08-30
Design Arena UI
1151 Elo2026-08-31
Design Arena UI
1151 Elo2026-09-01
Design Arena UI
1151 Elo2026-09-02
Design Arena UI
1149 Elo2026-09-03
Design Arena UI
1147 Elo2026-09-04
Design Arena Website
1164 Elo2026-07-28
Design Arena Website
1164 Elo2026-07-30
Design Arena Website
1164 Elo2026-07-27
Design Arena Website
1164 Elo2026-07-29
Design Arena Website
1164 Elo2026-07-31
Design Arena Website
1163 Elo2026-08-02
Design Arena Website
1163 Elo2026-08-01
Design Arena Website
1163 Elo2026-08-05
Design Arena Website
1163 Elo2026-08-03
Design Arena Website
1162 Elo2026-08-09
Design Arena Website
1163 Elo2026-08-04
Design Arena Website
1159 Elo2026-08-14
Design Arena Website
1163 Elo2026-08-06
Design Arena Website
1163 Elo2026-08-07
Design Arena Website
1162 Elo2026-08-10
Design Arena Website
1162 Elo2026-08-08
Design Arena Website
1150 Elo2026-08-18
Design Arena Website
1162 Elo2026-08-11
Design Arena Website
1150 Elo2026-08-15
Design Arena Website
1161 Elo2026-08-12
Design Arena Website
1159 Elo2026-08-13
Design Arena Website
1150 Elo2026-08-19
Design Arena Website
1149 Elo2026-08-16
Design Arena Website
1150 Elo2026-08-20
Design Arena Website
1159 Elo2026-08-27
Design Arena Website
1150 Elo2026-08-17
Design Arena Website
1155 Elo2026-09-05
Design Arena Website
1159 Elo2026-08-21
Design Arena Website
1159 Elo2026-08-22
Design Arena Website
1159 Elo2026-08-23
Design Arena Website
1159 Elo2026-08-24
Design Arena Website
1159 Elo2026-08-28
Design Arena Website
1160 Elo2026-08-25
Design Arena Website
1160 Elo2026-08-26
Design Arena Website
1152 Elo2026-09-06
Design Arena Website
1159 Elo2026-08-29
Design Arena Website
1158 Elo2026-08-30
Design Arena Website
1159 Elo2026-08-31
Design Arena Website
1159 Elo2026-09-01
Design Arena Website
1157 Elo2026-09-02
Design Arena Website
1156 Elo2026-09-03
Design Arena Website
1155 Elo2026-09-04
LMArena Creative Writing
1339 Elo2026-07-21
LMArena Creative Writing
1345 Elo2026-07-27
LMArena Creative Writing
1345 Elo2026-07-30
LMArena Creative Writing
1345 Elo2026-08-02
LMArena Creative Writing
1345 Elo2026-08-03
LMArena Creative Writing
1345 Elo2026-08-06
LMArena Creative Writing
1345 Elo2026-08-12
LMArena Creative Writing
1345 Elo2026-08-21
LMArena Creative Writing
1345 Elo2026-09-02
LMArena Creative Writing
1345 Elo2026-08-27
LMArena Creative Writing
1345 Elo2026-09-01
LMArena Long Query
1380 Elo2026-07-21
LMArena Long Query
1397 Elo2026-07-27
LMArena Long Query
1397 Elo2026-07-30
LMArena Long Query
1397 Elo2026-08-02
LMArena Long Query
1397 Elo2026-08-06
LMArena Long Query
1397 Elo2026-08-03
LMArena Long Query
1396 Elo2026-08-12
LMArena Long Query
1396 Elo2026-08-21
LMArena Long Query
1397 Elo2026-09-02
LMArena Long Query
1397 Elo2026-08-27
LMArena Long Query
1397 Elo2026-09-01
LMArena Math
1358 Elo2026-07-21
LMArena Math
1374 Elo2026-07-27
LMArena Math
1374 Elo2026-07-30
LMArena Math
1374 Elo2026-08-02
LMArena Math
1374 Elo2026-08-03
LMArena Math
1374 Elo2026-08-06
LMArena Math
1374 Elo2026-08-12
LMArena Math
1374 Elo2026-08-21
LMArena Math
1375 Elo2026-09-01
LMArena Math
1375 Elo2026-08-27
LMArena Text
1389 Elo2026-07-21
LMArena Text
1399 Elo2026-07-27
LMArena Text
1399 Elo2026-07-30
LMArena Text
1400 Elo2026-08-10
LMArena Text
1399 Elo2026-08-02
LMArena Text
1399 Elo2026-08-06
LMArena Text
1399 Elo2026-08-03
LMArena Text
1400 Elo2026-08-21
LMArena Text
1400 Elo2026-08-12
LMArena Text
1401 Elo2026-09-01
LMArena Text
1401 Elo2026-09-02
LMArena Text
1401 Elo2026-08-27
LMArena Vision
1175 Elo2026-07-12
LMArena Vision
1188 Elo2026-07-21
LMArena Vision
1189 Elo2026-07-26
LMArena Vision
1189 Elo2026-07-30
LMArena Vision
1189 Elo2026-08-02
LMArena Vision
1189 Elo2026-08-06
LMArena Vision
1189 Elo2026-08-21
LMArena Vision
1188 Elo2026-08-27
SWE-bench Verified
76.8%2025-09-02
SWE-bench Verified
76.8%2025-08-04
SWE-rebench
44.6% ± 0.62025-12-01tools

LMArena data licensed CC-BY-4.0. Aider (Apache-2.0) and SWE-bench (MIT) are open. Design Arena uses blind human preference. Submitted scaffold plus model result. Compare the harness and effort settings, not only the model name.

Task evidence

How broad is the support?

Each status is exact to this model version and the selected task. Official release claims remain provisional.

  • 70% of intended weighted task evidence across 3 independent sources. 6 of 6 configured evaluation feeds currently have results.

  • Best LLM for Business WritingProvisional evidence

    70% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for Marketing CopyProvisional evidence

    70% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for Chat & RoleplayProvisional evidence

    70% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 5 of 5 configured evaluation feeds currently have results.

  • 70% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for Creative WritingProvisional evidence

    60% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for WritingProvisional evidence

    60% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for Frontend & UIProvisional evidence

    60% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Coding or LMArena WebDev. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for CodingProvisional evidence

    56% of intended weighted task evidence across 4 independent sources; broader independent confirmation is still missing from LiveBench Coding or LMArena WebDev. 8 of 8 configured evaluation feeds currently have results.

  • Best Local LLM for CodingProvisional evidence

    56% of intended weighted task evidence across 2 independent sources; broader independent confirmation is still missing from LiveBench Coding or LMArena WebDev. 5 of 5 configured evaluation feeds currently have results.

  • 55% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Reasoning or VLMEvalKit tasks. 3 of 4 configured evaluation feeds currently have results.

  • 55% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Reasoning. 5 of 5 configured evaluation feeds currently have results.

  • 50% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Reasoning. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for Competition MathProvisional evidence

    40% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Mathematics or LiveBench Reasoning. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for Math & ReasoningProvisional evidence

    40% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Mathematics or LiveBench Reasoning. 5 of 5 configured evaluation feeds currently have results.

  • Best Value LLMProvisional evidence

    36% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or Route reliability. 3 of 4 configured evaluation feeds currently have results.

  • 33% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or Route reliability. 3 of 4 configured evaluation feeds currently have results.

  • Best LLM for Code CompletionProvisional evidence

    31% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Coding or LMArena WebDev. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for Autonomous AgentsProvisional evidence

    29% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Frontier-Bench or LiveBench Agentic Coding. 7 of 7 configured evaluation feeds currently have results.

  • Best LLM for Fast SummariesProvisional evidence

    29% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or Route reliability. 3 of 4 configured evaluation feeds currently have results.

  • 25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 4 of 5 configured evaluation feeds currently have results.

  • Best Cheap LLM for High VolumeProvisional evidence

    25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or Route reliability. 3 of 4 configured evaluation feeds currently have results.

  • 25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 4 of 5 configured evaluation feeds currently have results.

  • Cheapest Capable LLMProvisional evidence

    25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or Route reliability. 3 of 4 configured evaluation feeds currently have results.

  • Best LLM for TranslationProvisional evidence

    25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LiveBench Instruction Following or LiveBench Language. 4 of 5 configured evaluation feeds currently have results.

  • Best LLM for Massive ContextProvisional evidence

    25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or LongBench v2. 3 of 4 configured evaluation feeds currently have results.

  • Best LLM for OCR & DocumentsProvisional evidence

    25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or Structured-output eval. 3 of 5 configured evaluation feeds currently have results.

  • 25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or Structured-output eval. 3 of 5 configured evaluation feeds currently have results.

  • Best LLM for Charts & DiagramsProvisional evidence

    25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or Structured-output eval. 3 of 5 configured evaluation feeds currently have results.

  • Best LLM for Document ParsingProvisional evidence

    25% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from LMArena Document or Structured-output eval. 3 of 5 configured evaluation feeds currently have results.

  • Best LLM for Agentic CodingProvisional evidence

    23% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Design Arena Full-stack or Design Arena Web-apps. 8 of 8 configured evaluation feeds currently have results.

  • 20% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from FACTS Grounding or LiveBench Instruction Following. 4 of 6 configured evaluation feeds currently have results.

  • Best LLM for SummarizationProvisional evidence

    20% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from FACTS Grounding or LiveBench Instruction Following. 4 of 6 configured evaluation feeds currently have results.

  • Best LLM for AI AgentsProvisional evidence

    20% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LiveBench Agentic Coding. 7 of 7 configured evaluation feeds currently have results.

  • Best LLM for Hermes AgentProvisional evidence

    18% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for OpenHandsProvisional evidence

    18% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.

  • Best LLM for GooseProvisional evidence

    18% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.

  • 18% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from FACTS Grounding or LMArena Document. 4 of 6 configured evaluation feeds currently have results.

  • Best LLM for OpenClawProvisional evidence

    18% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 5 of 5 configured evaluation feeds currently have results.

  • 17% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LiveBench Data Analysis. 4 of 5 configured evaluation feeds currently have results.

  • 17% of intended weighted task evidence across 1 independent source; broader independent confirmation is still missing from Berkeley Function Calling or LiveBench Data Analysis. 4 of 5 configured evaluation feeds currently have results.

  • 13% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from LiveBench Data Analysis or Route reliability. 2 of 4 configured evaluation feeds currently have results.

  • 10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 4 of 5 configured evaluation feeds currently have results.

  • 10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from Berkeley Function Calling or LMArena Agent. 4 of 4 configured evaluation feeds currently have results.

  • Best LLM for Grounded RAGProvisional evidence

    10% of intended weighted task evidence across 0 independent sources; broader independent confirmation is still missing from CRAG or FACTS Grounding. 2 of 5 configured evaluation feeds currently have results.

Recommendation coverage

Where this model appears

Not currently included in a published recommendation. The task evidence panel above says exactly what is confirmed and what is still missing.

Catalog data via OpenRouter and models.dev. Last checked within the past 6 hours.