← All models

Claude Opus 4.8

Language Available Full comparison ↗

anthropic/claude-opus-4-8 · by Anthropic · decoder-only-transformer

Pricing — 1 offering(s)

Anthropic

Input tokens

  • $5.00 / 1M tokens (input) Current 2026-06-12 → present

    Opus 4 tier pricing (versions 4.5-4.8), unchanged since 2026-06-12.

Output tokens

  • $25.00 / 1M tokens (output) Current 2026-06-12 → present

    Opus 4 tier pricing (versions 4.5-4.8), unchanged since 2026-06-12.

Cache hits and refreshes

  • $0.50 / 1M tokens (cached input) Current 2026-06-12 → present

5-minute cache writes

  • $6.25 / 1M tokens (cache write) Current 2026-06-12 → present

Batch input tokens

  • $2.50 / 1M tokens (input) Current 2026-06-12 → present

Batch output tokens

  • $12.50 / 1M tokens (output) Current 2026-06-12 → present

Showing the active price and any recorded history. Full pricing history is available via the paid API — see API docs.

See how Claude Opus 4.8 fits into a cost-aware routing setup

See how →

Capability profile

How Claude Opus 4.8 rates across core capability dimensions, with the task-level evidence behind each rating.

reasoning Strong
coding Strong

SWE-bench Verified (500 human-verified GitHub issues; patch must pass the repo's own tests; no file-location hints): resolves 88.6% — fails roughly 1 issue in 9. Vendor-reported (Anthropic system card, averaged over 5 trials at max effort).

tool use Strong
instruction following Strong
context window Strong

1M-token context window at standard pricing — confirmed live against Anthropic's pricing docs (2026-07-20): 'Claude Fable 5, ... Claude Opus 4.8, Opus 4.7, Opus 4.6, Sonnet 5, and Sonnet 4.6 include the full 1M token context window at standard pricing.'

multilingual Strong
speed Weak

The largest/most capable tier — higher latency and cost than Sonnet/Haiku. A 'Fast mode' research preview exists at premium pricing ($10/$50 per MTok input/output) but is not the default.

cost efficiency Weak

Coding benchmarks

View the full coding benchmark leaderboard →
Benchmark Score Harness Source
swe-bench-verified ↗ 88.6% 2026-05 agentless 2.0 🏢 vendor ↗
swe-bench-pro 69.2% 2026-06 agentless 🏢 vendor ↗
terminal-bench-2-1 74.6% 2026-05 terminus-2 🏢 vendor ↗
terminal-bench-4-0 23.2% 2026-09 mini-swe-agent 📊 leaderboard ↗

Operator guidance

Route here for the most demanding reasoning/coding/agentic work when quality matters more than latency or price. Drop to Sonnet 5 for the usual price/performance balance, or Haiku 4 for high-volume, latency-sensitive work.

Use cases

Limitations

Citations