LLM Reference

Claude Opus 4.8

Researched today

Last refreshed 2026-05-28. Next refresh: weekly.

ProprietaryMultimodalCodingRAGAgentsLong contextVisionJSON / Tool useHighlight

Claude Opus 4.8 is worth evaluating for coding, rag, and agents when its provider route and context window match the workload.

Decision context: Coding task fit, 7 tracked provider routes, and research from 2026-05-28.

Use it for

  • Teams evaluating coding, rag, and agents
  • Workloads that can use a 1m context window
  • Buyers comparing 4 tracked provider routes

Do not use it for

  • Workloads where another current model has stronger sourced task evidence

Cheapest output

$25.00

Anthropic per 1M tokens

Provider routes

7

Tracked API hosts

Quality / dollar

Grade D

Ranked by benchmark score divided by cheapest output price

Freshness

2026-05-28

Researched today

fresh

Top use-case fit

Coding

Q/$ D

1 relevant benchmark in the decision map.

RAG

Included by capability and metadata signals in the decision map.

Agents

Included by capability and metadata signals in the decision map.

Provider price ladder

Compare all 7
ProviderInput / 1MOutput / 1MBatch in / outCacheRoute
Anthropic$5.00$25.00$2.50 / $12.50read $0.500 / 5m $6.25 / 1h $10.00
Serverless
AWS Bedrock$5.00$25.00--
Serverless
GCP Vertex AI$5.00$25.00--
Serverless
Microsoft Foundry$5.00$25.00--
Serverless

Benchmark peer barsfor Coding

Migration checks

No linked migration route is available for this model yet.

About

Claude Opus 4.8 is Anthropic's most capable generally-available model for complex reasoning, long-horizon agentic coding, and high-autonomy work. It improves on Opus 4.7 across agentic coding (SWE-bench Pro 69.2% vs 64.3%), agentic terminal coding (Terminal-Bench 2.1 74.6% vs 66.1%), multidisciplinary reasoning (HLE 57.9% with tools vs 54.7%), and agentic computer use (OSWorld-Verified 83.4%). Features a 1M-token context window (200k on Microsoft Foundry), 128k max output tokens, adaptive thinking with effort control defaulting to high, computer use, and a fast mode at 2.5× speed ($10/$50 per MTok). Approximately 4× less likely than Opus 4.7 to leave undetected code flaws. Deprecated Claude Opus 4 and Claude Sonnet 4 should migrate to Claude Opus 4.8.

Claude Opus 4.8 has a 1m-token context window.

Claude Opus 4.8 input tokens at $5/1M, output at $25/1M.

Capabilities

VisionMultimodalReasoningFunction CallingTool UseStructured OutputsCode ExecutionPrompt CachingBatch API

Benchmark Scores(6)

Scores are benchmark-specific and are direction-aware: the same numeric gap can mean very different outcomes across suites. Use the leaderboard context and this model's provider route to decide whether the winning margin is meaningful for your workload.
BenchmarkScoreVersionSource
SWE-bench Pro69.2SWE-bench Prohttps://www.anthropic.com/news/claude-opus-4-8
Terminal-Bench 2.174.6Terminal-Bench 2.1https://www.anthropic.com/news/claude-opus-4-8
Humanity's Last Exam57.9with toolshttps://www.anthropic.com/news/claude-opus-4-8
OSWorld83.4OSWorld-Verifiedhttps://www.anthropic.com/news/claude-opus-4-8
GDPval-AA1890.0https://www.anthropic.com/news/claude-opus-4-8
Finance Agent v253.9https://www.anthropic.com/news/claude-opus-4-8

Rankings

Specifications

Released2026-05-28
Context1m
Max output128,000
ArchitectureDecoder Only
Knowledge cutoff2026-01
Specializationgeneral
LicenseProprietary
Trainingfinetuned

Created by

Developing safe and ethical AI systems.

San Francisco, California, United States
Founded 2021
Website