Currently listed through these providers:
Model details
GPT-5 Chat (latest)
GPT-5 Chat (latest) is documented within OpenAI's developer API catalog under the identifier gpt-5-chat-latest, and OrcaRouter surfaces this model for application builders seeking the chat-tuned variant of the GPT-5 family. A community thread dated August 7, 2025 captures early hands-on experimentation with the GPT-5 family, where users compared latency against earlier generations. That contemporaneous signal lines up with the catalog's release timing and suggests practitioners were probing the model's responsiveness in real workflows as soon as it appeared in third-party routers. The practical fit for this model is conversational and reasoning-heavy use: developers can lean on it for natural multi-turn dialogue while still benefiting from multimodal inputs and structured response handling, which are surfaced as supported capabilities in the surrounding OpenAI documentation navigation. Because the available excerpts only confirm the documentation entry and a community discussion, the overview deliberately avoids claims about training scale, parameter counts, benchmark leadership, or architectural lineage that the sources do not directly state.
The model's intended role is to serve as a chat-optimized endpoint that complements the broader GPT-5 line, giving teams a dependable surface for assistant-style applications rather than a raw completion model. For builders choosing a router, the appeal is straightforward access to a chat-aligned variant from a familiar OpenAI documentation path, with the flexibility to pair it alongside reasoning, attachment handling, and structured output in product pipelines. Where the evidence stops, the writing stops: no specific training corpus, parameter count, benchmark score, or context-window behavior is asserted here, since none of those details appear in the supplied excerpts. Teams evaluating the model for production should rely on the catalog quick-info panel for numeric limits, pricing, and modality flags, and use this overview as a qualitative orientation to where the model sits in the GPT-5 family and how early adopters have begun exploring it.
Quick Info
Powered by- Provider
- OrcaRouter
- Model key
- openai/gpt-5-chat-latest
- Release date
- Aug 7, 2025
- Last updated
- Aug 7, 2025
- Knowledge cutoff
- 2024-09-30
- Input modalities
- Output modalities
- Capabilities
Cost
- Input token cost
- $1.25
- Output token cost
- $10.00
Limits
- Input tokens
- 272,000 tokens
- Output tokens
- 100,000 tokens
- Context window
- 400,000 tokens