GPT-4.1 logo

GPT-4.1

OpenAI · Released Apr 2025

Recommended

The default migration target for GPT-4 Turbo workloads: same request shape, lower cost, 1M-token context — convert legacy function_call payloads to the tools format and move on.

Is it right for you?

How it performs by task

Coding

Very Good

Strong instruction following on diffs and long files; the practical floor for production coding work

Agents

Very Good

Reliable tools-format function calling; legacy function_call payloads are rejected — convert before migrating

Long documents

Very Good

1M-token context handles full-corpus passes that needed chunking on GPT-4 Turbo

Pricing

Input

$2 / 1M tokens

Output

$8 / 1M tokens

Context

1M tokens

Verified 2026-06-12View full pricing

Benchmarks

BenchmarkScoreSource
MMLU90.2% Source
SWE-bench Verified54.6% Source