Claude Opus
Anthropic's line for long-running agentic coding and knowledge work.
Claude Opus is Anthropic's line for work that runs long and thinks hard: multi-step coding agents, changes across a whole repository, and knowledge work that wants the full context window. Opus 5.5, the current version, sits below Fable and above Sonnet in Anthropic's lineup, at a quarter of Fable's price.
What it is good at
- Long multi-step coding runs — Terminal-Bench 4.0 agentic coding at 66.4% (vendor-reported)
- Knowledge work with tools — Humanity's Last Exam 67.7% with tools (vendor-reported)
- Computer use — OSWorld 2.1 at 81.8% (vendor-reported)
What it is less good at
- Not the cheapest per token — Sonnet 5.5 is half the price, Haiku 5.5 a fraction of it
- Below Fable on the hardest long-horizon work; Anthropic positions Fable for when Opus at higher effort falls short
Reaching it with TeamCLI
TeamCLI runs Claude Opus as Claude Opus 5.5:
Ways to reach it
| Route | Spends | Status |
|---|---|---|
| Claude Code | An Anthropic subscription (Claude Pro or Max), or an Anthropic API key. | Verified |
| Cursor | A Cursor plan. | Listed — not yet verified |
| Antigravity | A Google account on a Google AI plan (the Antigravity / Google AI Ultra pools). | Listed — not yet verified |
The seat line
Each route maps to a TeamCLI seat line — the same fields, written generically. No machine paths, no launcher script of anybody's, no team's own seat names.
| Route | cli | vendor | model | version | display | launch |
|---|---|---|---|---|---|---|
| Claude Code | claude-code | anthropic | Claude Opus | 5.5 | Claude Opus 5.5 | claude |
| Cursor | cursor | anthropic | Claude Opus | 5.5 | Claude Opus 5.5 | cursor-agent |
| Antigravity | antigravity | anthropic | Claude Opus | 5.5 | Claude Opus 5.5 | agy |
Versions
One block per version, the current one first, each with its own anchor. The comparison table puts the costs and the family's benchmark columns side by side; the values are verbatim, and every one is sourced and dated in its version's section below and in the Sources list. A dash means the source publishes no figure, and nothing here is estimated.
No independent intelligence or speed index is republished here: Artificial Analysis measures both across labs, and its terms allow the link but not the copy — its figures live on its own pages.
| Version | Released | Context | Input price | Output price | Terminal-Bench 4.0 (agentic coding) | FrontierCode v1.1, Main (agentic coding) | CursorBench 4.0 (agentic coding) | Humanity's Last Exam (multidisciplinary reasoning) | OSWorld 2.1 (computer use) | Terminal-Bench-Science 0.1 (agentic science) | GDPval-AA v2.1 (knowledge work, Elo) | AutomationBench |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Claude Opus 5.5 — current | September 22, 2026 | 1M tokens | $4 / MTok | $20 / MTok | 66.4% | 54.4% | 57.8% | 67.7% (with tools) | 81.8% (partial) | 58.7% | 1846 | 40.0% |
| Claude Opus 5 | — | — | $5 / MTok | $25 / MTok | 52.3% | 48.0% | 46.6% | — | — | — | — | — |
| Claude Opus 4.8 | — | — | $5 / MTok | $25 / MTok | — | — | — | — | — | — | — | — |
| Claude Opus 4.7 | — | — | $5 / MTok | $25 / MTok | — | — | — | — | — | — | — | — |
| Claude Opus 4.6 | — | — | $5 / MTok | $25 / MTok | — | — | — | — | — | — | — | — |
| Claude Opus 4.5 | — | — | $5 / MTok | $25 / MTok | — | — | — | — | — | — | — | — |
Claude Opus 5.5 — current
- Model id:
claude-opus-5-5
Abilities
- Released: September 22, 2026 — Anthropic — Claude Opus 5.5 model page (read 2026-10-19).
- Context window: 1M tokens — Anthropic — Claude Opus 5.5 model page (read 2026-10-19).
- Max output: 128K tokens — Anthropic — Claude Opus 5.5 model page (read 2026-10-19).
- Output speed: —
- Latency: Moderate (comparative, within Anthropic's lineup) — Anthropic — models overview (read 2026-10-19).
Costs
- Input price: $4 / MTok — Anthropic — pricing (read 2026-10-19). Batch API 50% off; cache reads $0.20
- Output price: $20 / MTok — Anthropic — pricing (read 2026-10-19).
- Fast mode: $8 / MTok in, $40 / MTok out — Anthropic — Claude Opus 5.5 announcement (read 2026-10-19). Up to 2.5× the output speed
Benchmarks — as reported by the lab
- Terminal-Bench 4.0 (agentic coding) — 66.4% (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19) — xhigh effort; ±2.6 pts standard error
- FrontierCode v1.1, Main (agentic coding) — 54.4% (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19)
- CursorBench 4.0 (agentic coding) — 57.8% (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19)
- Humanity's Last Exam (multidisciplinary reasoning) — 67.7% (with tools) (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19)
- OSWorld 2.1 (computer use) — 81.8% (partial) (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19)
- Terminal-Bench-Science 0.1 (agentic science) — 58.7% (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19)
- GDPval-AA v2.1 (knowledge work, Elo) — 1846 (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19)
- AutomationBench — 40.0% (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19)
This is the version the seat lines above run.
Claude Opus 5
- Model id:
claude-opus-5
Abilities
- Released: —
- Context window: —
- Max output: —
- Output speed: —
- Latency: —
Costs
- Input price: $5 / MTok — Anthropic — pricing (read 2026-10-19).
- Output price: $25 / MTok — Anthropic — pricing (read 2026-10-19).
Benchmarks — as reported by the lab
- Terminal-Bench 4.0 (agentic coding) — 52.3% (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19)
- FrontierCode v1.1, Main (agentic coding) — 48.0% (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19)
- CursorBench 4.0 (agentic coding) — 46.6% (Anthropic — Claude Opus 5.5 announcement, read 2026-10-19)
Claude Opus 4.8
- Model id: —
Abilities
- Released: —
- Context window: —
- Max output: —
- Output speed: —
- Latency: —
Costs
- Input price: $5 / MTok — Anthropic — pricing (read 2026-10-19).
- Output price: $25 / MTok — Anthropic — pricing (read 2026-10-19).
Claude Opus 4.7
- Model id: —
Abilities
- Released: —
- Context window: —
- Max output: —
- Output speed: —
- Latency: —
Costs
- Input price: $5 / MTok — Anthropic — pricing (read 2026-10-19).
- Output price: $25 / MTok — Anthropic — pricing (read 2026-10-19).
Claude Opus 4.6
- Model id: —
Abilities
- Released: —
- Context window: —
- Max output: —
- Output speed: —
- Latency: —
Costs
- Input price: $5 / MTok — Anthropic — pricing (read 2026-10-19).
- Output price: $25 / MTok — Anthropic — pricing (read 2026-10-19).
Claude Opus 4.5
- Model id: —
Abilities
- Released: —
- Context window: —
- Max output: —
- Output speed: —
- Latency: —
Costs
- Input price: $5 / MTok — Anthropic — pricing (read 2026-10-19).
- Output price: $25 / MTok — Anthropic — pricing (read 2026-10-19).
Sources
- Anthropic — Claude Opus 5.5 model page — read 2026-10-19
- Anthropic — pricing — read 2026-10-19
- Anthropic — models overview — read 2026-10-19
- Anthropic — Claude Opus 5.5 announcement — read 2026-10-19
Figures refreshed 2026-10-19; benchmarks are vendor-reported. Prices are per million tokens unless a figure says otherwise.