ModelRefs / Claude Opus 4 vs Claude Sonnet 4 — Benchmarks, Pricing & Co…
Claude Opus 4 vs Claude Sonnet 4 — Benchmarks, Pricing & Co…
Claude Opus 4 vs Claude Sonnet 4: side-by-side benchmarks, pricing, context windows, coding ability and deployment. Pick the right model for your stack.
Overview
Claude Opus 4: Claude Opus 4 is Anthropic's May 2025 hybrid-reasoning model for complex coding, reasoning, multimodal analysis, and long-running agent workflows. The dated claude-opus-4-20250514 snapshot was retired on Anthropic-operated platforms on June 15, 2026 and is retained as a historical evaluation and migration reference.
Claude Sonnet 4: Claude Sonnet 4 is Anthropic's May 2025 hybrid-reasoning model for coding, visual reasoning, tool use, and agent workflows. The dated claude-sonnet-4-20250514 snapshot was retired on Anthropic-operated platforms on June 15, 2026 and is retained as a historical evaluation and migration reference.
Context window: Claude Opus 4 accepts up to 200,000 tokens, Claude Sonnet 4 up to 200,000. Confirm the limit for your specific deployment channel before relying on it.
Claude Opus 4 vs Claude Sonnet 4 at a glance
| Attribute | Claude Opus 4 | Claude Sonnet 4 |
|---|---|---|
| Provider | Anthropic | Anthropic |
| Released | 2025-05-22 | 2025-05-22 |
| Context window | 200,000 tokens | 200,000 tokens |
| Input price | $15/M tokens | $3/M tokens |
| Output price | $75/M tokens | $15/M tokens |
| Licence | Proprietary | Proprietary |
| Self-hostable | No, hosted only | No, hosted only |
| Modalities | text, vision | text, vision |
Benchmarks reported for both
| Benchmark | Claude Opus 4 | Claude Sonnet 4 | Reported |
|---|---|---|---|
| aime-2025 | 75.5 | 70.5 | 2025-05-22 |
| gpqa | 79.6 | 75.4 | 2025-05-22 |
| mmmu | 76.5 | 74.4 | 2025-05-22 |
| swe-bench | 72.5 | 72.7 | 2025-05-22 |
| tau-bench | 59.6 | 60 | 2025-05-22 |
| terminal-bench | 43.2 | 35.5 | 2025-05-22 |
Scores are as published by their cited sources on the dates shown. Different runs use different harnesses, prompts, and dates, so treat these as directional evidence rather than a head-to-head result.
Where they differ most
- Reasoning: Claude Opus 4 78%, Claude Sonnet 4 73%. Claude Opus 4 leads on this dimension.
- Coding: Claude Opus 4 60%, Claude Sonnet 4 57%. Claude Opus 4 leads on this dimension.
- Agentic: Claude Opus 4 63%, Claude Sonnet 4 61%. Claude Opus 4 leads on this dimension.
- Multimodal: Claude Opus 4 77%, Claude Sonnet 4 74%. Claude Opus 4 leads on this dimension.
- OCR: Claude Opus 4 77%, Claude Sonnet 4 74%. Claude Opus 4 leads on this dimension.
- Tool Use: Claude Opus 4 71%, Claude Sonnet 4 70%. Claude Opus 4 leads on this dimension.
Capability scores are ModelRefs' own derived signals, not vendor claims or benchmark results. Validate against your own workload before relying on them.
Continue your research
Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Claude Opus 4 vs Claude Sonnet 4 — Benchmarks, Pricing & Co….
Frequently asked questions
Which is better, Claude Opus 4 or Claude Sonnet 4?
Claude Opus 4 and Claude Sonnet 4 target different workloads — see the benchmark and pricing tables for a side-by-side answer.
Is Claude Opus 4 cheaper than Claude Sonnet 4?
Compare $0.0150 vs $0.00300 on this page.