anthropic

Anthropic: Claude Sonnet 4.6 (batch)

The Anthropic Claude Sonnet 4.6 is a large language model designed for various input modalities, including text, image, and file inputs, with a context length of 1 million tokens. It supports tools and reasoning capabilities, allowing for more complex processing tasks. When considering this model, note that it has a high price point: $1.5 per million input tokens and $7.5 per million output tokens. Its benchmark standing is also worth examination: a blended score of 71.3 across three independent benchmarks suggests solid performance but may not be exceptional in all areas. Those with specific needs for image or file inputs, or who value the tool and reasoning support, may find this model worthwhile despite its price; others might prefer alternatives with lower costs or more comprehensive benchmark coverage.

Quality Score
100/100
price + capability + benchmarks
Input Price
$1.50
per 1M tokens · checked 2026-09-23
Output Price
$7.50
per 1M tokens · checked 2026-09-23
Context Window
1,000,000
tokens

Benchmark results

Independent, published benchmarks. Blended score 69.7 across 3 benchmarks, last refreshed 2026-09-23. How scoring works →

BenchmarkMeasuresScore
Aider Polyglot code editing across languages 61.3
AI Index broad capability composite 49.6
EQ-Bench emotional understanding in dialogue 80.0
Model ID
anthropic/claude-sonnet-4.6:batch
Vendor
anthropic
Released
February 2026
Tokenizer
Claude
Input Modalities
text, image, file
Output Modalities
text
Max Output
128,000 tokens
Tool Calling
✓ supported
Structured Output
✓ supported
Reasoning Mode
✓ supported
Vision
✓ accepts images
Audio
no
Moderated
yes

What it costs in practice

Computed from the current $1.50/M input and $7.50/M output rates. Run your own numbers →

JobTokensCost
Summarize a 50-page report 30k in / 1.5k out $0.06
Classify 1,000 customer emails 500k in / 50k out $1.12
A month of a busy support chatbot 5M in / 2M out $22.50

Strong choice for

Category rankings

Where Anthropic: Claude Sonnet 4.6 (batch) places across the 60 categories it ranks in. How we rank →

#CategoryScore
#2 SQL GenerationCode · of 25 ranked 155
#2 Code DocumentationCode · of 25 ranked 144
#2 CSV / Spreadsheet CleanupData · of 25 ranked 146
#2 Long-Document SummarizationWriting · of 25 ranked 162
#2 Blog Post WritingWriting · of 25 ranked 138
#2 Transcript CleanupWriting · of 25 ranked 146
#2 Character RoleplayPersonal · of 25 ranked 150
#2 Trip PlanningPersonal · of 25 ranked 138
#3 Code RefactoringCode · of 25 ranked 162
#3 Unit Test GenerationCode · of 25 ranked 144
#3 Regex WritingCode · of 25 ranked 131
#3 CI/CD PipelinesCode · of 25 ranked 138
#3 Frontend Component DesignCode · of 25 ranked 137
#3 ETL ScriptingData · of 25 ranked 150
#3 OCR / Document ParsingData · of 25 ranked 140
#3 Table Extraction from PDFsData · of 25 ranked 140
#3 Marketing CopyWriting · of 25 ranked 126
#3 Language TranslationWriting · of 25 ranked 131
#3 Medical Note SummarizationProfessional · of 25 ranked 150
#3 Physics TutoringEducation · of 25 ranked 143
#3 History TutoringEducation · of 25 ranked 145
#3 Standardized Test PrepEducation · of 25 ranked 126
#3 Meeting NotesBusiness · of 25 ranked 147
#3 Resume WritingBusiness · of 25 ranked 115
#3 RFP ResponseBusiness · of 25 ranked 168
#3 Screenshot DebuggingVision · of 25 ranked 137
#3 Chart & Graph ReadingVision · of 25 ranked 143
#3 RAG PipelinesAgents · of 25 ranked 162
#3 Long-Context Q&AAgents · of 25 ranked 157
#3 Tutoring for KidsPersonal · of 25 ranked 131
#4 Code ReviewCode · of 25 ranked 161
#4 Bug FixingCode · of 25 ranked 166
#4 Creative WritingWriting · of 25 ranked 155
#4 Screenwriting & DialogWriting · of 25 ranked 155
#4 Academic WritingWriting · of 25 ranked 155
#4 Legal DraftingProfessional · of 25 ranked 155
#4 Legal ResearchProfessional · of 25 ranked 167
#4 Contract ReviewProfessional · of 25 ranked 173
#4 Scientific ResearchProfessional · of 25 ranked 167
#4 Essay GradingEducation · of 25 ranked 161
#4 Sales / Cold EmailBusiness · of 25 ranked 120
#4 Job Application DraftingBusiness · of 25 ranked 120
#4 OKRs & Strategic PlanningBusiness · of 25 ranked 155
#4 Agent WorkflowsAgents · of 25 ranked 171
#4 Coding AgentsAgents · of 25 ranked 171
#4 Fiction CollaboratorPersonal · of 25 ranked 167
#4 Journaling HelperPersonal · of 25 ranked 120
#4 Recipe GenerationPersonal · of 25 ranked 120
#4 Math ProofsResearch · of 25 ranked 155
#4 Scientific CodingResearch · of 25 ranked 166
#4 Literature ReviewResearch · of 25 ranked 167
#5 Data AnalysisData · of 25 ranked 147
#5 Financial AnalysisProfessional · of 25 ranked 142
#5 Math TutoringEducation · of 25 ranked 137
#5 Browser AutomationAgents · of 25 ranked 147
#5 Fitness CoachingPersonal · of 25 ranked 126
#5 Experiment DesignResearch · of 25 ranked 143
#7 SOP / Process DocsBusiness · of 25 ranked 127
#7 Diagram ExtractionVision · of 25 ranked 138
#12 Function / Tool CallingAgents · of 25 ranked 127

Similar models

Quick answers

How much does Anthropic: Claude Sonnet 4.6 (batch) cost?
$1.50 per million input tokens and $7.50 per million output tokens.
What is Anthropic: Claude Sonnet 4.6 (batch)'s context window?
1,000,000 tokens, roughly 1,500 pages of text in a single request.
Does Anthropic: Claude Sonnet 4.6 (batch) support tool calling?
It supports tool calling, structured output, a reasoning mode.
Can Anthropic: Claude Sonnet 4.6 (batch) process images?
Yes, it accepts image input.

The Model Movers Report

One email every Friday, built from this site's own rankings: the current top five by benchmark score, every model released in the last seven days, and one note worked out from that week's numbers. You can unsubscribe from any issue with one click.