Claude Opus 5
Anthropic released 2026-07-24 proprietary
Nothing is both better and cheaper.
Under a balanced workload, no other model in the catalogue scores higher and costs less. This model is on the value frontier.
The same model via batch is 50% cheaper ($5.00/M) — same weights, different latency.
Our take
editorial — not a measurementThe best price/intelligence point at the frontier right now and the model most teams should default to for hard agentic work. It tops the Artificial Analysis intelligence index at 63 while costing $5/$25 — half of Fable 5, which scores lower. The flat 1M-token context with no tiered pricing is a genuine structural advantage over OpenAI, Google and xAI, all of which roughly double their rate past a threshold; a 900k-token Opus 5 request costs the same per token as a 9k one, which nothing else at this tier can say. Caveats: ~41s TTFT at max effort, and the same ~30% tokenizer inflation as Fable 5, so compare it to older models on cost-per-task, not cost-per-token.
- Unverified in this take: This note cites a time-to-first-token figure that the data layer withheld as implausible (it appears to be total reasoning time mislabelled at source). Treat it as unverified.
- Unverified in this take: Anthropic does publish a token-inflation figure, but it compares newer Claude models against EARLIER CLAUDE models — not against another vendor. The cross-vendor version of this claim is unsupported, so read any comparison to a non-Claude price with that in mind.
Strengths
- Highest Artificial Analysis intelligence index observed (63)
- Flat pricing across the full 1M window — no long-context cliff
- Half the price of Fable 5 at equal-or-better measured intelligence
- Effort parameter (low through max) gives real cost/quality control
- Fast-mode variant available at 2x for latency-sensitive work
Weaknesses
- ~41s time-to-first-token at max effort
- ~30% more tokens for the same text than pre-4.7 models
- Fast mode is first-party API only and excluded from batch
Reach for it when
- Complex agentic coding
- Long-context document and codebase work
- Default frontier choice for enterprise workloads
Avoid it if
- Sub-second first token is required
- The task is simple enough for Sonnet 5
Sources: platform.claude.complatform.claude.comartificialanalysis.ai
Every price dimension
| Input | $5.00/M |
|---|---|
| Output | $25.00/M |
| Cached input | $0.500/M |
| Cache write | $6.25/M |
| Cache write 1h | $10.00/M |
| Web search | $0.01/call |
| Batch discount | 50% |
Independent scores
| Intelligence | 63.1 |
|---|---|
| Coding | 78 |
| Agentic | 59.2 |
| LMArena Elo | 1504.2high |
Sources are listed separately rather than averaged. Across the 62 models both have scored they correlate at r = 0.806 — close agreement overall, but four models rank very differently between them, and a blended score would hide exactly those.
Best rankings by task
- svg #1of 98 1368
- codecategories #2of 145 1350
- uicomponent #2of 140 1371
- 3d #2of 136 1380
- website #3of 151 1332
- gamedev #3of 144 1381
- fullstack #3of 59 1353
- dataviz #4of 142 1358
Capability
| Context window | 1M |
|---|---|
| Max output | 128K |
| Input modes | text, image, file |
| Tool use | yes |
| Reasoning | optional |
| Open weights | no |
Provenance
| Price source | openrouter.ai |
|---|---|
| Fetched | 2026-08-24 |
| Quality data | verified |
| Cross-checked | vendor page |
A published time-to-first-token figure for this model was implausible (41s) and has been withheld rather than displayed.
Claude Opus 5 is Anthropic’s flagship model for demanding reasoning, coding, and long-horizon agentic work. It is particularly strong at end-to-end software tasks, code review and bug finding, visual analysis...