Apr 15 • General Discussion 💬
OpenAI Just Released GPT-4.1 to the API
It outperforms GPT-4o across key benchmarks, including:
- Coding: 54.6% on SWE-bench Verified (vs 33.2% for GPT-4o)
- Instruction Following: 38.3% on Multi Challenge (vs 27.8%)
- Long Context: Handles up to 1 million tokens with improved retrieval and reasoning
There are three versions now available:
- GPT-4.1: Full version with best performance
- GPT-4.1 mini: Matches or exceeds GPT-4o with 83% lower cost and lower latency
- GPT-4.1 nano: Fastest and lowest cost, ideal for classification and autocomplete tasks
Key improvements:
- Better accuracy in long document analysis and multi-step reasoning
- More reliable instruction following and formatting
- Reduced extraneous edits in code generation
- Enhanced ability to build functional AI agents
- Long context available at no additional cost
- Prompt caching discount increased to 75%
Cost (per 1M tokens):
- GPT-4.1 – $2 input / $8 output (blended: ~$1.84)
- GPT-4.1 mini – $0.40 input / $1.60 output (blended: ~$0.42)
- GPT-4.1 nano – $0.10 input / $0.40 output (blended: ~$0.12)
GPT-4.1 is only available via API at the moment.
(GPT-4.5 will be gone on July 14, 2025)
4
4 comments
OpenAI Just Released GPT-4.1 to the API
skool.com/ai-automation-society
A community built to master no-code AI automations. Join to learn, discuss, and build the systems that will shape the future of work.
Leaderboard (30-day)
1
🔥
+2486
2
🔥
+1133
3
+1055
4
🔥
+888
5
🔥
+722
Powered by