OpenAI: GPT-4.1
Overview
openai/gpt-4.1
GPT-4.1 is a flagship large language model optimized for advanced instruction following, real-world software engineering, and long-context reasoning. It supports a 1 million token context window and outperforms GPT-4o and GPT-4.5 across coding (54.6% SWE-bench Verified), instruction compliance (87.4% IFEval), and multimodal understanding benchmarks. It is tuned for precise code diffs, agent reliability, and high recall in large document contexts, making it ideal for agents, IDE tooling, and enterprise knowledge retrieval.
by OpenAI
| Input type | Output type | Publish time |
|---|---|---|
| 2025-04-14 |
Recent activity on GPT-4.1
Tokens processed per day
Throughput
| Providers | Min (tokens/s) | Max (tokens/s) | Avg (tokens/s) |
|---|---|---|---|
| Azure | 25.6 | 73.3 | 50.65 |
| OpenAI | 22.11 | 69.45 | 50.61 |
First Token Latency
| Providers | Min (ms) | Max (ms) | Avg (ms) |
|---|---|---|---|
| Azure | 453 | 1690 | 988.03 |
| OpenAI | 591 | 1334 | 1188.28 |
Uptime stats for GPT-4.1
Uptime stats for GPT-4.1 across all providers
| Uptime | Providers for GPT-4.1 |
|---|---|
| Azure | OpenAI |
Azure
Latency
1.55
s
Throughput
84.77
tps
Uptime
100.00
%
Recent uptime
Apr 30, 2026 6:00 PM-
Price
| Type | Price |
|---|---|
| Input | $ 2/ M tokens |
| Output | $ 8/ M tokens |
| Cache read | $ 0.5/ M tokens |
| Web search | $ 0.014/ counts |
Model limitation
- Context: 1.05M
- Max output: 32.77K
Supported Parameters
- max_completion_tokens
- temperature
- top_p
- frequency_penalty
- presence_penalty
- seed
- logit_bias
- logprobs
- top_logprobs
- response_format
- stop
- tools
- tool_choice
- parallel_tool_calls
Model Protocol Compatibility
- OpenAI Chat Completions
- OpenAI Responses
- OpenAI Embeddings
- Anthropic Messages
- Google Gemini
- Google Imagen
- Google Video
Sample code and API for GPT-4.1
ZenMux normalizes requests and responses across providers for you.
from openai import OpenAI
client = OpenAI(
base_url="https://zenmux.ai/api/v1",
api_key="<ZENMUX_API_KEY>",
)
# Chat Completion
completion = client.chat.completions.create(
model="openai/gpt-4.1",
messages=[\
{\
"role": "user",\
"content": "What is the meaning of life?"\
}\
]
)
print(completion.choices[0].message.content)