Quick answer
Use Opus 5.5 as the default Claude model for production coding. Increase effort first, then escalate to Fable 5.1 only when the task still needs more capability.
- Released
- September 22, 2026
- API model
- claude-opus-5-5
- Price
- $4 in / $20 out per MTok; $0.20/MTok cache reads; Fast mode costs $8/$40
- Context
- 1M tokens (128K max output)
01
Claude Opus 5.5: verified facts
| Field | Verified value |
|---|---|
| Vendor | Anthropic |
| Release | September 22, 2026 |
| API identifier | claude-opus-5-5 |
| Input / output price | $4 in / $20 out per MTok |
| Cache / fast pricing | $0.20/MTok cache reads; Fast mode costs $8/$40 |
| Context / max output | 1M tokens (128K max output) |
| Availability | Current recommended default; medium effort by default |
Every number on this page was checked against the linked vendor documentation on September 23, 2026. Prices are API list prices per million tokens; product subscriptions and regional availability are separate.
02
Where this model fits
Anthropic says Opus 5.5 costs 40% less per typical workload and produces output more than 30% faster than Opus 5. The base API price also falls from $5/$25 to $4/$20. Adaptive thinking is always enabled.
Your AI workspace for shipping software.
Download AIDEN free and point it at your existing Claude Code or Codex setup. No credit card, running in minutes.
Download AIDEN freeFree to start · macOS 12+ · No credit card required
03
When to use it
Use case 1
Default Claude model for production coding
Use case 2
Long-running migrations and repository-wide audits
Use case 3
Knowledge work that needs strong quality without Fable pricing
| Strengths | Trade-offs |
|---|---|
| Anthropic's recommended starting point for most workloads | Fable 5.1 remains the escalation path for the hardest reasoning tasks |
| 1M-token context and 128K maximum output at $4/$20 per MTok | Adaptive thinking is always on and cannot be disabled |
| More than 30% faster output than Opus 5 in Anthropic's tests | Fast mode doubles token prices |
04
Published benchmarks
| Benchmark | Score | Source / caveat |
|---|---|---|
| Terminal-Bench 4.0 | 66.4% | Anthropic launch report, Vendor-published at xhigh effort |
| FrontierCode v1.1 | 54.4% | Anthropic launch report, Vendor-published main-set result |
| GDPval-AA v2.1 | 1846 Elo | Anthropic launch report, Vendor-published result |
05