Claude Opus 5 Complete Guide: Features, Pricing, and Availability
Everything you need to know about Claude Opus 5, launched July 24, 2026: benchmarks, the new effort toggle, 1M context window, pricing, and how it fits into the Claude 5 family.
TL;DR
Claude Opus 5, released on July 24, 2026, delivers near-frontier intelligence at half the cost of Claude Fable 5. It scores 96.0% on SWE-bench Verified and 79.2% on SWE-bench Pro, ships with a 1M-token context window, and introduces a per-request effort toggle (low/medium/high). Pricing is unchanged from Opus 4.8 at $5 per million input tokens and $25 per million output tokens. It is the new default model on Claude Max and the strongest model available on Claude Pro.
What Claude Opus 5 Is
Opus is Anthropic's workhorse frontier tier: the model most teams reach for when a task needs real reasoning but doesn't justify Fable-class pricing. Claude Opus 5 keeps that role and closes most of the gap to the very top of the lineup. Anthropic describes it as "approaching Fable-level performance at half the cost," and the benchmark results back that up.
Opus 5 is the fourth model Anthropic has shipped in under two months, following Mythos 5, Fable 5, and Sonnet 5 in June 2026. That cadence signals a deliberate strategy: cover every point on the cost-capability curve with a current-generation model.
Key Specifications
- Model ID: claude-opus-5
- Context window: 1M tokens (default and maximum)
- Max output: 128K tokens standard; 300K via the Message Batches API (beta)
- Extended thinking: on by default
- Effort control: low / medium / high, set per request
- Pricing: $5/M input, $25/M output; fast mode at $10/$50 runs about 2.5× faster
Benchmark Highlights
- SWE-bench Verified: 96.0%
- SWE-bench Pro: 79.2% — within a point of Mythos 5 (80.3%) and Fable 5 (80.0%), and far ahead of Opus 4.8's 69.2%
- Frontier-Bench v0.1 (agentic coding): 43.3%, ahead of Fable 5 (33.7%) and GPT-5.6 Sol (34.4%)
- ARC-AGI-3 (novel reasoning): 30.2%, roughly three times the next-best model
- GDPval-AA v2 (knowledge work, Elo): 1,861 vs Fable 5's 1,747
- OSWorld 2.0 (computer use): surpasses Fable 5 at one-third the cost
Availability
Claude Opus 5 is available immediately on claude.ai, in Claude Code, in Claude Cowork, and through the Claude API. On consumer plans it becomes the default model for Claude Max subscribers and the strongest model available to Claude Pro subscribers.
How It Fits the Claude 5 Family
- Fable 5: still the pick for long-horizon agents that run autonomously for days
- Mythos 5: leads on frontier cybersecurity and demanding biology research, restricted to approved organizations
- Sonnet 5: best for high-volume, well-scoped tasks
- Opus 5: the new default for everything in between — serious coding, research, and agentic work at half of Fable pricing
Conclusion
Claude Opus 5 resets the price-performance baseline for frontier AI. If you are running Opus 4.8 today, the upgrade is free in dollar terms and roughly doubles agentic coding performance. If you are paying for Fable 5 on tasks that finish within a session, Opus 5 will likely do the same work at half the cost.