Model Details
GPT-4.1 mini
Quick answer: GPT-4.1 mini is OpenAI's cost-efficient mid-tier model from the GPT-4.1 family, released April 14, 2025 alongside GPT-4.1 and GPT-4.1 nano. It scores 31.8% on BigCodeBench and 0.0 on SHADE-Arena. At $0.40/$1.60 per 1M tokens with a 1M-token context window, it is the most capable OpenAI model in the $0.40 input price range.
At a Glance
Where GPT-4.1 mini leads
- 1M-token context window — 8× the context of GPT-4o mini (128K), at $0.40 input
- $0.40/$1.60 per 1M — cost-efficient with full 1M context
- 31.8% BigCodeBench — competitive general coding at the mini tier
- Multimodal vision support
- GPT-4.1 training lineage — improved instruction following vs GPT-4o mini
Where it lags
- 0.0 SHADE-Arena — low agentic sabotage detection (indicates safe baseline but limited agentic reasoning)
- Below GPT-4.1 (32.8% BigCodeBench) — expected for the mini tier
- No reasoning mode
- June 2024 knowledge cutoff
Best for: High-volume pipelines requiring 1M context at $0.40 input — document analysis, whole-repository code tasks, and large-batch inference that would be too expensive at GPT-4.1 pricing ($2/$8).
What GPT-4.1 mini Is
GPT-4.1 mini is the mid-tier model in the GPT-4.1 family, sitting between GPT-4.1 ($2/$8) and GPT-4.1 nano (not yet shown) on the capability/cost spectrum. Its headline feature is the 1M-token context window at $0.40 input — making it the most cost-efficient way to access 1M context on the OpenAI API.
Released alongside GPT-4.1 on April 14, 2025, GPT-4.1 mini replaced GPT-4o mini as the recommended cost-tier model for most workloads, offering the same price range ($0.40 vs $0.15 input) with dramatically more context (1M vs 128K tokens) and better overall performance from the GPT-4.1 training lineage.
For budget-constrained workloads that don't require 1M context, Gemini 2.5 Flash at $0.15/$0.60 provides better benchmark coverage. For workloads that specifically benefit from 1M context at minimal cost, GPT-4.1 mini is the strongest proprietary option.
Specifications
| Field | Value |
|---|---|
| Organization | OpenAI |
| Parameters | Undisclosed |
| Context window | 1,000,000 tokens |
| License | Proprietary (API only) |
| Release date | April 14, 2025 |
| Knowledge cutoff | June 2024 |
| Modality | Text + Vision (multimodal) |
Pricing
| Input (per 1M tokens) | Output (per 1M tokens) | |
|---|---|---|
| OpenAI API | $0.40 | $1.60 |
Prompt caching provides a 75% discount on cached input. Pricing per OpenAI pricing page.
Context Window
GPT-4.1 mini has a 1,000,000-token context window — the same as GPT-4.1 and Gemini 2.5 Pro, at a fraction of their cost.
Public Benchmark Scores
| Benchmark | Score | Source | Date |
|---|---|---|---|
| BigCodeBench | 31.8% | Benchgen evaluation | 2025-07 |
| SHADE-Arena | 0.0 overall success | Anthropic research post | 2025-06 |
GPT-4.1 mini vs Alternatives
| Model | Context | BigCodeBench | Price (in/out per 1M) |
|---|---|---|---|
| GPT-4.1 mini | 1M | 31.8% | $0.40 / $1.60 |
| GPT-4.1 | 1M | 32.8% | $2 / $8 |
| GPT-4o mini | 128K | 27.5%* | $0.15 / $0.60 |
| Gemini 2.5 Flash | 1M | — | $0.15 / $0.60 |
*GPT-4o mini LiveCodeBench; BigCodeBench not directly compared. GPT-4.1 mini vs GPT-4o mini: 8× more context, better training, at 2.7× higher input cost.
Related Models & Benchmarks
- Compare: GPT-4.1 · GPT-4o mini · Gemini 2.5 Flash
- Benchmarks: BigCodeBench · SHADE-Arena
Frequently Asked Questions
What is GPT-4.1 mini?
GPT-4.1 mini is OpenAI's April 2025 mid-tier model with a 1M-token context window at $0.40/$1.60 per 1M tokens. It scores 31.8% BigCodeBench and supports vision input.What is GPT-4.1 mini's context window?
GPT-4.1 mini supports a 1,000,000-token context window — the same as the full GPT-4.1 model.How much does GPT-4.1 mini cost?
$0.40 per 1M input tokens and $1.60 per 1M output tokens, with a 75% cached input discount.Should I use GPT-4.1 mini or GPT-4o mini?
For tasks requiring long context (>128K tokens), GPT-4.1 mini is the clear choice — 1M vs 128K context at 2.7× higher input cost. For tasks that fit within 128K tokens and are highly cost-sensitive, GPT-4o mini or Gemini 2.5 Flash are better value.Specs and scores sourced from OpenAI's official GPT-4.1 announcement (April 14, 2025) and Benchgen evaluations. Pricing cited to the OpenAI pricing page. Last updated 2026-07-24.
Benchmark Leaderboards
This model isn’t on any benchmark leaderboard yet.