openai·chat· vision

GPT 5.4

openai/gpt-5.4

OpenAI's new flagship — agentic workloads and deep reasoning chains.

At a glance

  • · $2.50 input / $15.00 output per 1M tokens (provider list price, billed in credits)
  • · Context window: 400,000 tokens
  • · OpenAI-compatible — model: "openai/gpt-5.4" (provider: openai)
  • · Best for: Better tool-use accuracy than GPT-5.3 with shorter outputs

Strengths

  • +Better tool-use accuracy than GPT-5.3 with shorter outputs
  • +Native vision for screenshots, charts, and documents
  • +400k context — fits code repos and long conversations

Weaknesses

  • Pricier per token than Sonnet 4.6 at comparable coding quality
  • Reasoning mode noticeably increases latency over standard calls

Use cases

  • Agentic pipelines with multiple tool calls per turn
  • Long-context analysis: read a whole repo, find one bug
  • Structured extraction from PDFs and screenshots

Not the right pick for

  • ×Bulk classification at scale — too pricey, use gpt-5.4-mini

Alternatives in the catalog

How to call it

Drop-in OpenAI-compatible. Just swap the `model` string and you're using this model.

curl https://www.getmorecredits.com/v1/chat/completions \
  -H "Authorization: Bearer $GMC_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "openai/gpt-5.4",
    "messages": [{ "role": "user", "content": "Hello!" }]
  }'

Good for

Frequently asked questions

How much does GPT 5.4 cost?

Input is $2.50 per million tokens; output is $15.00 per million tokens — provider list price. On getmorecredits.com you pay in credits, no subscription, no minimum spend.

What's the context window of GPT 5.4?

GPT 5.4 accepts up to 400,000 input tokens, with a maximum of 16,384 output tokens per response. That fits whole code repos or book-length documents into a single call.

What can GPT 5.4 do?

GPT 5.4 supports image input (vision), tool calling, native JSON mode, structured output, reasoning mode. OpenAI's new flagship — agentic workloads and deep reasoning chains.

Is GPT 5.4 OpenAI-compatible?

Yes. On getmorecredits.com you call GPT 5.4 with the same request shape as OpenAI: POST /v1/chat/completions with "model": "openai/gpt-5.4". Existing OpenAI SDKs (Python, Node, etc.) work by changing the base URL and API key.