OpenAI
MultimodalPriced via OpenRouter

GPT-4.1 (batch)

text+image+file->text

Long-context coding-focused GPT-4 successor with a 1M-token window. Aimed at developers who need predictable instruction following over very long inputs.

GPT-4.1 (batch) is a multimodal AI model from OpenAI. It costs $1.00 per million input tokens and $4.00 per million output tokens (blended $1.90/M), with a 1,047,576-token context window.

Profile inherited from upstream GPT-4.1 ↗ — this is a hosted variant of the same open-weights model.

RatesUSD per million tokens. Blend is 70% input, 30% output, our calculation.

INPUT
$1.00/M
per million input tokens
OUTPUT
$4.00/M
per million output tokens
BLENDED 70/30
$1.90/M
unchanged since 3 May
CONTEXT
1,047,576
tokens

From OpenRouter's listing, read on 17 Nov 2026.

What it is good at
  • 1M context window
  • Strong coding
  • Better instruction following than GPT-4o
  • Cheap relative to GPT-4
Typical use cases
  • Repo-scale coding
  • Long document analysis
  • Cost-sensitive vision

Benchmarks

vs. best public score
Scores inherited from GPT-4.1 — this is a hosted variant of the same open-weights model, so the underlying benchmark scores are identical.
MMLU90.2%
Multitask academic knowledge across 57 subjects. Source: OpenAI (self-reported), MMLU, read 2026-11-10.
Graduate-level science questions, "Google-proof". Source: OpenAI (self-reported), GPQA diamond (no tools), read 2026-11-10.
HumanEval94.5%
Python function synthesis from docstrings. Source: OpenAI (self-reported), HumanEval, read 2026-11-10.
Real GitHub issues solved end-to-end. Source: OpenAI (self-reported), SWE-bench Verified, read 2026-11-10.
LMArena Elo1415 Elo
Crowd-sourced head-to-head preference Elo rating. Source: LMArena (independent), LMArena Text Arena score, read 2026-11-10.
Hand-curated from each provider's published reports and public leaderboards. Methodology varies across sources, treat as directional rather than authoritative.

Also reported by the provider

Results on benchmarks outside the set we compare across models. They are listed for completeness and are deliberately excluded from the bars above and from every ranking on this site, because a score is only comparable against the same benchmark variant. Why benchmark scores do not compare →

  • MATH-50082.1%self-reported by OpenAI · 2025-07 · gpt-4.1 row, assistant prompt. simple-evals README, zero-shot chain of thought. Its footnote says models from o1 onward are scored on MATH-500, not the full MATH test set. Not comparable with the MATH column, so shown here and never ranked.

More from OpenAI

See all 28 →

Frequently asked questions

How much does GPT-4.1 (batch) cost?

GPT-4.1 (batch) costs $1.00 per million input tokens and $4.00 per million output tokens, for a blended reference rate of $1.90 per million tokens.

What is GPT-4.1 (batch)'s context window?

GPT-4.1 (batch) supports up to 1,047,576 tokens of context in a single request.

What is GPT-4.1 (batch) best for?

GPT-4.1 (batch) is well suited to 1M context window, Strong coding and Better instruction following than GPT-4o.

Who makes GPT-4.1 (batch)?

GPT-4.1 (batch) is developed and served by OpenAI. It was released in Apr 2025.

Terms used on this page