← All models

Claude Opus 4

Language Retired Full comparison ↗

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

Pricing — 1 offering(s)

Anthropic

Input tokens

Retired — no longer sold · last price $15.00 / 1M tokens (input), until 2026-06-14

  • $15.00 / 1M tokens (input) Historical 2026-06-12 → 2026-06-14 Correction
    Corrected 2026-09-25: Modelglass first recorded a wrong price for these dates; this row replaces it, and the…

    Corrected 2026-09-25: Modelglass first recorded a wrong price for these dates; this row replaces it, and the wrong row is no longer served.

Output tokens

Retired — no longer sold · last price $75.00 / 1M tokens (output), until 2026-06-14

  • $75.00 / 1M tokens (output) Historical 2026-06-12 → 2026-06-14 Correction
    Corrected 2026-09-25: Modelglass first recorded a wrong price for these dates; this row replaces it, and the…

    Corrected 2026-09-25: Modelglass first recorded a wrong price for these dates; this row replaces it, and the wrong row is no longer served.

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 fits into a cost-aware routing setup

See how →

Capability profile

How Claude Opus 4 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 72.5% — fails roughly 1 issue in 4. Vendor-reported (Anthropic, Claude Opus 4 launch; agentless scaffold).

tool use Strong
instruction following Strong
multilingual Strong
speed Weak

The largest/most capable tier — higher latency and cost than Sonnet/Haiku.

cost efficiency Weak

Coding benchmarks

View the full coding benchmark leaderboard →
Benchmark Score Harness Source
swe-bench-verified ↗ 72.5% 2025-05 agentless 2.0 🏢 vendor ↗
aider-polyglot ↗ 72.2% 2025-06 aider 📊 leaderboard ↗
livecodebench ↗ 57.8% 2025-05 — 📊 leaderboard ↗

Operator guidance

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

Use cases

Limitations

Citations