High-volume, cost-sensitive work
Choose Claude Opus 4.7
Claude Opus 4.7 costs 11% less on a typical mix of three input tokens for every output token ($10.00 against $11.25 per million tokens).
Choose Chat Latest for the latest release. Choose Claude Opus 4.7 for high-volume, cost-sensitive work, writing and code generation and very long inputs.
Each answer below comes straight from the published prices, specifications and independent benchmark results.
Choose Claude Opus 4.7
Claude Opus 4.7 costs 11% less on a typical mix of three input tokens for every output token ($10.00 against $11.25 per million tokens).
Choose Claude Opus 4.7
Claude Opus 4.7 charges $25.00 per million output tokens, against $30.00 for Chat Latest. Output dominates the bill when answers are long.
Choose Claude Opus 4.7
Claude Opus 4.7 reads up to 1M tokens per request, against 400k for Chat Latest.
Choose Chat Latest
Chat Latest is the newer release (May 5, 2026, against April 2026 for Claude Opus 4.7).
| Volume | Chat Latest | Claude Opus 4.7 | Difference |
|---|---|---|---|
| Prototype: 1M input, 250K output tokens | $12.50 | $11.25 | Claude Opus 4.7, $1.25 less |
| Production: 20M input, 5M output tokens | $250.00 | $225.00 | Claude Opus 4.7, $25.00 less |
| Scale: 500M input, 100M output tokens | $5,500 | $5,000 | Claude Opus 4.7, $500.00 less |
Standard rates, without caching or batch discounts. Volumes are per month.
| Chat Latest | Claude Opus 4.7 | |
|---|---|---|
| Provider | OpenAI | Anthropic |
| Category | Language models | Language models |
| Status | Generally available | Legacy |
| Headline price | $11.25 / 1M tokens | $10.00 / 1M tokens |
| Input | $5.00 / 1M tokens | $5.00 / 1M tokens |
| Output | $30.00 / 1M tokens | $25.00 / 1M tokens |
| Cached input | $0.50 / 1M tokens | $0.50 / 1M tokens |
| Context window | 400k | 1M |
| Max output | 128k | 128k |
| Released | May 5, 2026 | April 2026 |
| Knowledge cutoff | April 2026 | No data |
| Licence | Proprietary | Proprietary |
| Capabilities | Vision, Tool calling, Structured output, Prompt caching, Batch API | Vision, Reasoning, Tool calling, Structured output, Prompt caching, Batch API |
| API model id | chat-latest | claude-opus-4-7 |
Leader in bold.
| Benchmark | Chat Latest | Claude Opus 4.7 |
|---|---|---|
| Terminal-Bench Hard | No data | 51.5% |
| GPQA Diamond | No data | 91.4% |
| Humanity's Last Exam | No data | 42.3% |
| IFBench (instruction following) | No data | 58.6% |
| τ²-Bench (agentic tool use) | No data | 88.6% |
| MMMU-Pro (visual reasoning) | No data | 78.8% |
| Long-context reasoning (AA-LCR) | No data | 78.7% |
Independent evaluations by Artificial Analysis, run the same way for every model, at the reasoning setting shown. Percentages unless marked as an index.
Choose Chat Latest for the latest release. Choose Claude Opus 4.7 for high-volume, cost-sensitive work, writing and code generation and very long inputs.
Claude Opus 4.7 is 11% cheaper than Chat Latest at the headline price ($10.00 / 1M tokens vs $11.25 / 1M tokens).
Claude Opus 4.7 reads more per request: 1M tokens against 400k.
Chat Latest is made by OpenAI and Claude Opus 4.7 by Anthropic. Claude Opus 4.7 is 11% cheaper than Chat Latest at the headline price ($10.00 / 1M tokens vs $11.25 / 1M tokens). Claude Opus 4.7 reads more per request: 1M tokens against 400k. Chat Latest is the newer model (May 5, 2026, against April 2026). Only Claude Opus 4.7 offers reasoning.
At 20 million input and 5 million output tokens a month, Chat Latest costs $250.00 and Claude Opus 4.7 costs $225.00, at standard rates.
Chat Latest was released on May 5, 2026 and Claude Opus 4.7 on April 2026.