Opus 4.7 vs GPT-5 for agents
Both of these have been superseded, and one of them we cannot serve at all. Here is the comparison people still search for, and the two models that actually replace them.
GPT-5 is not callable on llmrelay — requests return model_not_found. Opus 4.7 still runs but Opus 5 has superseded it. This page is a reference, not a recommendation to buy either. The current equivalent on llmrelay is GPT-5.6 Sol at $2.5/M input.
Callable on llmrelay: GPT-5 returns model_not_found on our upstream. GPT-5.6 Sol is the live current-generation option.
Prices are llmrelay's, at 50% of official list. Specs are the vendor's own published figures (Anthropic model docs), not our benchmarks. We do not publish scores we cannot source.
If you specifically need Opus 4.7
- +Reproducing output you already validated against this pinned snapshot.
- +Otherwise Opus 5 is the current pick at the same $2.50/M input, with a May 2026 knowledge cutoff.
If you came here for GPT-5
- +Not callable here, so we have nothing to sell you on this one.
- +GPT-5.6 Sol is the current-generation option on our key.
- +Running a Claude model and a GPT model against the same prompts is the cheapest way to settle which suits your agent — one key covers both.
The honest answer
Neither is the current answer. Opus 5 against GPT-5.6 Sol is the live version of this question, and both sit on one key at 50% off list, so you can test rather than speculate.
What this costs you per month
At 50M input and 5M output tokens a month — a realistic heavy agent workload.
Struck-through column is the vendor's list price for the same traffic. Adjust the numbers on the calculator.
One key covers both
Switching between them is a model-id change, not a new account. Prepaid, no subscription, $10 minimum — enough to run the same prompt through both and settle the argument with your own output.
Get API key →