OpenAI
GPT-4o Mini
gpt-4o-miniContext
128K
Input / 1M tokens
$0.1500
Output / 1M tokens
$0.6000
Weighted tokens formula
1× provider × 1× tier
About this model
GPT-4o Mini
GPT-4o Mini — the cheap, fast OpenAI tier. Strong for most tasks where you don't need Opus/Sonnet quality. The standard pairing for routers that escalate hard requests to a larger model.
Released
2024-07
Training cutoff
2023-10
Parameters
Not disclosed
Best for
Use cases
- ✓High-volume chat where cost matters
- ✓Classification + extraction
- ✓Routing in front of larger models
- ✓Cheap-tier customer support
Strengths
What it does well
- ●Sub-second TTFT
- ●Same tool calling reliability as GPT-4o
- ●128K context window despite small price
- ●Strong default for OpenAI-shaped workflows
Limitations
Know the trade-offs
- !Reasoning + code trails Claude Haiku 4.5 / Qwen 2.5 Coder
- !More hallucination than larger models
- !No vision — must use GPT-4o for multimodal
Benchmarks
Published scores
Scores from official model cards and public leaderboards. Higher is better unless noted.
| Benchmark | Score | Measures |
|---|---|---|
| MMLU | 82.0 | General knowledge across 57 subjects |
| HumanEval | 87.2 | Python code generation, pass@1 |
| MATH | 70.2 | Mathematics, mixed difficulty |
| MGSM | 87.0 | Multilingual grade school math |
Code example
Chat completion with GPT-4o Mini
from openai import OpenAI
client = OpenAI(
base_url="https://api.anoman.io/v1",
api_key="anm-sk-..."
)
response = client.chat.completions.create(
model="gpt-4o-mini",
messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)Weighted tokens
weighted_tokens = raw_tokens × 1 (provider) × 1 (tier)
Pro plan: 20M weighted tokens/month. Combined multiplier 1×: 20,000,000 raw tokens available.
Alternatives
Similar models in our catalog
Workloads that fit GPT-4o Mini often fit these too. Compare benchmarks and pricing before locking in a default.