token.app › Compare › GPT-5 vs MiniMax M3
GPT-5 vs MiniMax M3
GPT-5 and MiniMax M3 split the 2 shared benchmarks evenly. MiniMax M3 is 6.5× cheaper per token on a blended 3:1 basis. MiniMax M3 accepts the larger context window (1.0M).
Head to head
Blended price uses a 3:1 input:output mix
| GPT-5 | MiniMax M3 | |
|---|---|---|
| Provider | OpenAI | MiniMax |
| Input $/1M | $1.25 | $0.300 |
| Output $/1M | $10.00 | $1.20 |
| Blended $/1M | $3.44 | $0.525 |
| Context window | 400K | 1.0M |
| Max output | 128K | 512K |
| Released | Aug 2025 | May 2026 |
| Licence | Proprietary | Open weights |
| GPQA Diamond | 86.2% | 90.9% |
| OTIS Mock AIME | 91.4% | 71.1% |
Benchmark scores independently run and published by Epoch AI, 'AI Benchmarking Hub', used under CC BY 4.0. Only benchmarks with a score for BOTH models are compared.
Choose GPT-5 if…
- Higher OTIS Mock AIME (91.4% vs 71.1%)
Choose MiniMax M3 if…
- Costs less per token — $0.525 vs $3.44 blended
- Higher GPQA Diamond (90.9% vs 86.2%)
- Larger context window (1.0M)
- Open weights — self-hostable
- Newer model (released May 2026)
FAQ
Which is better, GPT-5 or MiniMax M3?
Across the 2 benchmarks both models have independently-run scores for, GPT-5 leads on 1 and MiniMax M3 leads on 1. Benchmarks are one input — price and context window are on this page too.
Is GPT-5 cheaper than MiniMax M3?
MiniMax M3 is cheaper: $0.525 versus $3.44 per million tokens blended at 3:1 input:output. Input alone: $1.25 vs $0.300. Output alone: $10.00 vs $1.20.
What context windows do GPT-5 and MiniMax M3 have?
GPT-5: 400,000 tokens. MiniMax M3: 1,048,576 tokens.
Who makes GPT-5 and MiniMax M3?
GPT-5 is made by OpenAI. MiniMax M3 is made by MiniMax.