GPT-5.4 mini
GPT-5.4 mini language model.
One model, automatic routing. Use the model ID in every request.
Model pricing
30% offEach line has its own published sales rates. Automatic routing chooses the supply used; fallback can change the price charged. Use the same model ID in every request. No credit is reserved; admitted requests may overdraw the balance, blocking new requests.
Line: 5cc03d79-0f38-441a-9717-6d0ef661ea4bInput / Output$0.525 / $3.15per 1,000,000 tokensPrice conditions
Line ID: 5cc03d79-0f38-441a-9717-6d0ef661ea4b
InputOfficial price: $0.75OTRouter price: $0.525
Cache writeOTRouter price: $0.525
Cached inputOTRouter price: $0.0525
OutputOfficial price: $4.50OTRouter price: $3.15
per 1,000,000 tokens
Standard official API priceOpenAI ↗
Rate effective from:
Protocols and limits
These are published baseline text paths. Routing prefers an applicable native path and also considers cooldowns and each request’s features. This configuration does not show the path chosen for a request or current line health.
Chat Completions / Chat CompletionsNative
Baseline request forms: JSON · Stream
Native features describe the entry declarations; they do not indicate live verification.
Native features are not declared for this path.
- Function tools
- Not declared
- Serial tool control
- Not declared
- Reasoning
- Not declared
- Image input
- Not declared
- Stop sequences
- Not declared
- JSON object output
- Not declared
- JSON schema output
- Not declared
- Streaming
- Not declared
Declared entry limits
- Context window
- 400,000 tokens
- Max input
- 400,000 tokens
- Max output
- 128,000 tokens
Effective routing limits
Effective limits combine the model and this entry. An undeclared entry limit remains unknown.
- Context window
- 400,000 tokens
- Max input
- 400,000 tokens
- Max output
- 128,000 tokens
Responses / ResponsesNative
Baseline request forms: JSON · Stream
Native features describe the entry declarations; they do not indicate live verification.
Native features are not declared for this path.
- Function tools
- Not declared
- Serial tool control
- Not declared
- Reasoning
- Not declared
- Image input
- Not declared
- Stop sequences
- Not declared
- JSON object output
- Not declared
- JSON schema output
- Not declared
- Streaming
- Not declared
Declared entry limits
- Context window
- 400,000 tokens
- Max input
- 400,000 tokens
- Max output
- 128,000 tokens
Effective routing limits
Effective limits combine the model and this entry. An undeclared entry limit remains unknown.
- Context window
- 400,000 tokens
- Max input
- 400,000 tokens
- Max output
- 128,000 tokens
Messages / MessagesNative
Baseline request forms: JSON · Stream
Native features describe the entry declarations; they do not indicate live verification.
Native features are not declared for this path.
- Function tools
- Not declared
- Serial tool control
- Not declared
- Reasoning
- Not declared
- Image input
- Not declared
- Stop sequences
- Not declared
- JSON object output
- Not declared
- JSON schema output
- Not declared
- Streaming
- Not declared
- Messages token counting
- Not declared
Declared entry limits
- Context window
- 400,000 tokens
- Max input
- 400,000 tokens
- Max output
- 128,000 tokens
Effective routing limits
Effective limits combine the model and this entry. An undeclared entry limit remains unknown.
- Context window
- 400,000 tokens
- Max input
- 400,000 tokens
- Max output
- 128,000 tokens
Line: e2a16f79-2e66-475d-b901-72eca9a3de17Input / Output$0.525 / $3.15per 1,000,000 tokensPrice conditions
Line ID: e2a16f79-2e66-475d-b901-72eca9a3de17
InputOfficial price: $0.75OTRouter price: $0.525
Cache writeOTRouter price: $0.525
Cached inputOTRouter price: $0.0525
OutputOfficial price: $4.50OTRouter price: $3.15
per 1,000,000 tokens
Standard official API priceOpenAI ↗
Rate effective from:
Protocols and limits
These are published baseline text paths. Routing prefers an applicable native path and also considers cooldowns and each request’s features. This configuration does not show the path chosen for a request or current line health.
Chat Completions / Chat CompletionsNative
Baseline request forms: JSON · Stream
Native features describe the entry declarations; they do not indicate live verification.
- Function tools
- Supported
- Serial tool control
- Not declared
- Reasoning
- Not declared
- Image input
- Not declared
- Stop sequences
- Not declared
- JSON object output
- Not declared
- JSON schema output
- Not declared
- Streaming
- Supported
Declared entry limits
- Context window
- 1,050,000 tokens
- Max input
- 1,050,000 tokens
- Max output
- 128,000 tokens
Effective routing limits
Effective limits combine the model and this entry. An undeclared entry limit remains unknown.
- Context window
- 400,000 tokens
- Max input
- 400,000 tokens
- Max output
- 128,000 tokens
Responses / ResponsesNative
Baseline request forms: JSON · Stream
Native features describe the entry declarations; they do not indicate live verification.
- Function tools
- Supported
- Serial tool control
- Not declared
- Reasoning
- Not declared
- Image input
- Not declared
- Stop sequences
- Not declared
- JSON object output
- Not declared
- JSON schema output
- Not declared
- Streaming
- Supported
Declared entry limits
- Context window
- 1,050,000 tokens
- Max input
- 1,050,000 tokens
- Max output
- 128,000 tokens
Effective routing limits
Effective limits combine the model and this entry. An undeclared entry limit remains unknown.
- Context window
- 400,000 tokens
- Max input
- 400,000 tokens
- Max output
- 128,000 tokens
Messages / MessagesNative
Baseline request forms: JSON · Stream
Native features describe the entry declarations; they do not indicate live verification.
- Function tools
- Supported
- Serial tool control
- Not declared
- Reasoning
- Not declared
- Image input
- Not declared
- Stop sequences
- Not declared
- JSON object output
- Not declared
- JSON schema output
- Not declared
- Streaming
- Supported
- Messages token counting
- Not declared
Declared entry limits
- Context window
- 1,050,000 tokens
- Max input
- 1,050,000 tokens
- Max output
- 128,000 tokens
Effective routing limits
Effective limits combine the model and this entry. An undeclared entry limit remains unknown.
- Context window
- 400,000 tokens
- Max input
- 400,000 tokens
- Max output
- 128,000 tokens
Line observations
UnavailablePublic line observations are not available yet. Missing values do not indicate a healthy or failed line.
- Historical delivery success
- —
- Completion latency
- —
- Published snapshot
- —
Historical observations are not an SLA. Completion latency is not time to first token (TTFT).
Request examples
Use the same model ID; no line parameter is required.
curl https://router.omnitools.ai/v1/chat/completions \
-H "Authorization: Bearer $OTRouter_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.4-mini",
"messages": [
{
"role": "user",
"content": "Hello, please introduce yourself."
}
],
"max_completion_tokens": 160
}'Model capabilities
- Router request protocols
- Chat CompletionsResponsesMessages
- Routing configuration
- Configured routes: 2Configured upstream systems: 1No baseline HA
Capabilities and limits depend on the line and protocol path. Review the published configuration with each line’s price.
These are Router request protocols. Routing prefers an applicable native line; otherwise, compatible conversion may be used. Requests whose meaning cannot be preserved are rejected. This list does not establish native upstream support.
Baseline HA means plain-text JSON and streaming requests have cross-provider safe-fallback candidates on Chat, Responses, and Messages. It is not live health or an SLA, and does not cover extra capabilities such as tools or stop sequences.
Model company: OpenAI
Protocol documentation