O4 Mini
Lightweight reasoning model from OpenAI supporting text and image inputs with a 200,000-token context window and 100,000 max output tokens. It is tuned for STEM problem solving with a June 2024 knowledge cutoff.
Certified & verified by nRouter· facts checked 2026-09-14 against 2 official sources
OpenAI
OpenAI
Benchmarks
Pricing
Rates are read live
Pricing for o4-mini is fetched from the live catalogue on load, so it is never served from a cache that could outlive a repricing.
Pricing Transparency & Zero-Markup Guarantee
nRouter strictly operates under Rule #28: Zero per-token markup. All inference requests for O4 Mini are billed at exact upstream provider list prices with zero per-token margin, zero routing fees, and zero hidden platform overhead.
Token consumption is settled at the published upstream rate of the exact served model and provider tier that executed your call.
Upstream prompt caching discounts (up to 50%–90% reduction on cached prompt tokens) pass through directly to your balance with zero added fee.
Every response includes the authoritative x-nr-request-cost header for instant accounting and FinOps reconciliation.
Specifications
reasoning
Context window, compared
Bars are scaled to 400K tokens. A model with no published context window reads Unknown, never zero.
Context Limits & Caching Economics
With an effective context ceiling of 200.0K tokens and a maximum output generation ceiling of 100K tokens, o4-mini is engineered to support long-document analysis, repository code exploration, and complex multi-turn dialog without token truncation.
Upstream prefix caching stores static context (system prompts, tool definitions, documentation corpora) in memory. Cached token reads receive up to 50%–90% cost discounts from upstream providers, passed through directly to your nRouter balance with zero added fee.
- ✦Autonomous Tool Use: Structured JSON schema validation and multi-step function calling loops.
- ✦Code Intelligence: Multi-file codebase refactoring, unit test generation, and complex algorithmic reasoning.
- ✦High-Density RAG: In-context information extraction across large token windows without needle loss.
- ✦Multimodal Ingestion: Visual documentation analysis, image parsing, and structured asset extraction.
Drop-in Integration
import os
from openai import OpenAI
# Drop-in replacement: point OpenAI client to nRouter gateway
client = OpenAI(
base_url="https://api.nrouter.ai/v1",
api_key=os.environ.get("NROUTER_API_KEY"), # Your nRouter virtual key (sk-nrouter-*)
)
response = client.chat.completions.create(
model="openai/o4-mini", # Exact model ID or smart router alias (e.g. nrouter/auto)
messages=[
{"role": "system", "content": "You are an expert engineer."},
{"role": "user", "content": "Explain multi-region failover and distributed consensus."}
],
temperature=0.7, # Sampling randomness (0.0 = deterministic, 1.0 = creative)
max_tokens=2048, # Upper ceiling on tokens generated
stream=False, # Set True to receive streaming Server-Sent Events (SSE)
)
print(response.choices[0].message.content)- base_url / baseURL
- Target endpoint for the proxy gateway. Use
https://api.nrouter.ai/v1for OpenAI-compatible tools orhttps://api.nrouter.aifor Anthropic Messages requests. - api_key / NROUTER_API_KEY
- Your nRouter virtual key (
sk-nrouter-*). Implements hardware-backed tenant isolation, model ACLs, per-key token budgets, and sub-millisecond preflight verification. - model
- Exact model identifier (
openai/o4-mini) or a dynamic smart router alias (such asnrouter/auto) for automated multi-provider intent tiering. - temperature & max_tokens
temperaturetunes sampling variance (0.0 for deterministic code/schema tasks, 0.7+ for creative generation);max_tokensenforces a hard ceiling on completion length to eliminate runaway compute spend.
Multi-Provider Redundancy
nRouter routes traffic for O4 Mini across multiple underlying cloud hyperscalers to guarantee continuous zero-downtime availability and eliminate vendor lock-in:
Traffic for openai/o4-mini routes dynamically across Microsoft Azure, AWS Bedrock, Google Cloud Vertex AI, and direct provider endpoints without code changes.
If an upstream cloud provider reports rate limits (HTTP 429), regional degradation, or an outage, nRouter automatically retries on an alternate cloud backend in <1ms.
A single nRouter virtual key provides unified access with enforced spending budgets, model allowlists, and end-to-end SOC 2 compliant audit logging.
Availability
Not enough data yet
We have not collected enough health probes for this model to publish an uptime or latency figure. A percentage from a handful of samples is fabricated precision, so none is shown until the sample count supports one.
Interactive Playground
Send a real test request to o4-mini using your virtual key.
Test it
curl https://api.nrouter.ai/v1/chat/completions \
-H "Authorization: Bearer $NROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{"model":"openai/o4-mini","messages":[{"role":"user","content":"Hello!"}]}'