Anthropic: Claude Opus 4.8
The Anthropic Claude Opus 4.8 is a large language model that supports three input modalities: text, image, and file. It has a maximum context length of 1 million tokens and can handle tasks requiring tools, reasoning, and output formatting. Those with a budget of over $5 per million input tokens may want to consider shortlisting this model for its impressive benchmark scores across four independent tests, including a blended score of 94.2. Its pricing is not the lowest among models with similar capabilities, but its performance standing makes it a viable option for users who require high-quality output and are willing to pay the premium.
Benchmark results
Independent, published benchmarks. Blended score 94.2 across 4 benchmarks, last refreshed 2026-09-04. How scoring works →
| Benchmark | Measures | Score |
|---|---|---|
| AI Index | broad capability composite | 94.6 |
| AI Index Coding | software engineering tasks | 100.0 |
| AI Index Agentic | multi-step tool-using tasks | 81.5 |
| EQ-Bench | emotional understanding in dialogue | 83.3 |
- Model ID
- anthropic/claude-opus-4.8
- Vendor
- anthropic
- Released
- May 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 $5.00/M input and $25.00/M output rates. Run your own numbers →
| Job | Tokens | Cost |
|---|---|---|
| Summarize a 50-page report | 30k in / 1.5k out | $0.19 |
| Classify 1,000 customer emails | 500k in / 50k out | $3.75 |
| A month of a busy support chatbot | 5M in / 2M out | $75.00 |
Category rankings
Where Anthropic: Claude Opus 4.8 places across the 29 categories it ranks in. How we rank →
| # | Category | Score |
|---|---|---|
| #11 | Bug FixingCode · of 25 ranked | 188 |
| #11 | Scientific CodingResearch · of 25 ranked | 188 |
| #12 | Code ReviewCode · of 25 ranked | 174 |
| #12 | Code RefactoringCode · of 25 ranked | 171 |
| #12 | ETL ScriptingData · of 25 ranked | 159 |
| #12 | Creative WritingWriting · of 25 ranked | 164 |
| #12 | Screenwriting & DialogWriting · of 25 ranked | 164 |
| #12 | Academic WritingWriting · of 25 ranked | 164 |
| #12 | Legal DraftingProfessional · of 25 ranked | 164 |
| #12 | Legal ResearchProfessional · of 25 ranked | 176 |
| #12 | Contract ReviewProfessional · of 25 ranked | 186 |
| #12 | Medical Note SummarizationProfessional · of 25 ranked | 159 |
| #12 | Scientific ResearchProfessional · of 25 ranked | 176 |
| #12 | Physics TutoringEducation · of 25 ranked | 151 |
| #12 | History TutoringEducation · of 25 ranked | 149 |
| #12 | Essay GradingEducation · of 25 ranked | 174 |
| #12 | RFP ResponseBusiness · of 25 ranked | 181 |
| #12 | OKRs & Strategic PlanningBusiness · of 25 ranked | 164 |
| #12 | RAG PipelinesAgents · of 25 ranked | 171 |
| #12 | Long-Context Q&AAgents · of 25 ranked | 161 |
| #12 | Fiction CollaboratorPersonal · of 25 ranked | 176 |
| #12 | Math ProofsResearch · of 25 ranked | 164 |
| #12 | Literature ReviewResearch · of 25 ranked | 176 |
| #14 | Long-Document SummarizationWriting · of 25 ranked | 164 |
| #14 | Agent WorkflowsAgents · of 25 ranked | 200 |
| #14 | Coding AgentsAgents · of 25 ranked | 200 |
| #20 | Character RoleplayPersonal · of 25 ranked | 152 |
| #22 | Unit Test GenerationCode · of 25 ranked | 156 |
| #22 | CI/CD PipelinesCode · of 25 ranked | 146 |
Similar models
Anthropic: Claude Fable 5.1 (batch)
Claude Opus 5
Anthropic: Claude Fable 5 (batch)
Anthropic: Claude Opus 4.7
Anthropic: Claude Opus 4.6
Anthropic: Claude Opus 4.5
Quick answers
- How much does Anthropic: Claude Opus 4.8 cost?
- $5.00 per million input tokens and $25.00 per million output tokens.
- What is Anthropic: Claude Opus 4.8's context window?
- 1,000,000 tokens, roughly 1,500 pages of text in a single request.
- Does Anthropic: Claude Opus 4.8 support tool calling?
- It supports tool calling, structured output, a reasoning mode.
- Can Anthropic: Claude Opus 4.8 process images?
- Yes, it accepts image input.