Best MiniMax M2.5 Alternatives
MiniMax M2.5 by MiniMax is a open source model priced at $0.3/1.2 per 1M tokens (in/out). It's already affordable, but you might want different strengths or features.
MiniMax M2.5
MiniMaxOpen SourceInput
$0.3/1M
Output
$1.2/1M
Context
200K
Max Output
128K
Why Switch from MiniMax M2.5?
Top Alternatives
Dramatically cheaper (72% less), open-source and self-hostable.
Input
$0.14/1M
53% cheaper
Output
$0.28/1M
77% cheaper
Context
164K
Max Output
16K
23% cheaper, comparable performance, 1M context (5x more).
Input
$0.31/1M
3% more
Output
$0.85/1M
29% cheaper
Context
1M
Max Output
32K
46% cheaper, 10M context (50x more), adds vision.
Input
$0.18/1M
40% cheaper
Output
$0.63/1M
48% cheaper
Context
10M
Max Output
32K
Reasoning tier option.
Input
$0.55/1M
83% more
Output
$2.19/1M
83% more
Context
128K
Max Output
64K
50% cheaper, 1M context (5x more), adds vision, tool-use.
Input
$0.15/1M
50% cheaper
Output
$0.6/1M
50% cheaper
Context
1M
Max Output
66K
Comparable performance, adds vision, tool-use.
Input
$0.4/1M
33% more
Output
$1.6/1M
33% more
Context
200K
Max Output
100K
Comparable performance, adds vision.
Input
$0.6/1M
100% more
Output
$2.2/1M
83% more
Context
200K
Max Output
128K
Comparable performance, 1M context (5x more), adds vision, tool-use.
Input
$0.5/1M
67% more
Output
$3/1M
150% more
Context
1M
Max Output
66K
Full Comparison Table
| Model | Input $/1M | Output $/1M | Context | MMLU-Pro | HumanEval | Score |
|---|---|---|---|---|---|---|
| DeepSeek V3DeepSeek | $0.1453% cheaper | $0.2877% cheaper | 164K | 78%-4.0% | 89%-1.0% | 83 |
| Llama 4 MaverickMeta | $0.313% more | $0.8529% cheaper | 1M | 80.5%-1.5% | 90.2%+0.2% | 78 |
| Llama 4 ScoutMeta | $0.1840% cheaper | $0.6348% cheaper | 10M | 74.2%-7.8% | 86%-4.0% | 70 |
| DeepSeek R1DeepSeek | $0.5583% more | $2.1983% more | 128K | 84%+2.0% | 92%+2.0% | 68 |
| Gemini 2.5 FlashGoogle | $0.1550% cheaper | $0.6050% cheaper | 1M | 76%-6.0% | 89.5%-0.5% | 66 |
| o3OpenAI | $0.4033% more | $1.6033% more | 200K | 87%+5.0% | 94.5%+4.5% | 63 |
| GLM-4.7Zhipu AI | $0.60100% more | $2.2083% more | 200K | 84.3%+2.3% | — | 61 |
| Gemini 3 FlashGoogle | $0.5067% more | $3.00150% more | 1M | 78%-4.0% | 90%Same | 56 |
| Mistral Large 3Mistral | $2.00567% more | $5.00317% more | 128K | 83%+1.0% | 91%+1.0% | 53 |
| GPT-4o MiniOpenAI | $0.1550% cheaper | $0.6050% cheaper | 128K | 68%-14.0% | 87.2%-2.8% | 52 |
| Gemini 3.1 ProGoogle | $2.00567% more | $12.00900% more | 1M | 91%+9.0% | 95%+5.0% | 50 |
| Gemini 3 ProGoogle | $2.00567% more | $12.00900% more | 1M | 89.8%+7.8% | 94%+4.0% | 50 |
| Mistral Medium 3Mistral | $0.4033% more | $2.0067% more | 128K | 76%-6.0% | 87%-3.0% | 50 |
| GLM-5Zhipu AI | $1.00233% more | $3.20167% more | 200K | 70.4%-11.6% | 91%+1.0% | 48 |
| o4-miniOpenAI | $1.10267% more | $4.40267% more | 200K | 85%+3.0% | 93.5%+3.5% | 46 |
| Claude Opus 4.6Anthropic | $5.001567% more | $25.001983% more | 200K | 89.5%+7.5% | 95%+5.0% | 43 |
| GPT-5.3 CodexOpenAI | $2.00567% more | $16.001233% more | 200K | 90%+8.0% | 96.5%+6.5% | 43 |
| GPT-5.2 CodexOpenAI | $1.75483% more | $14.001067% more | 200K | 89%+7.0% | 95.5%+5.5% | 43 |
| GPT-5OpenAI | $1.25317% more | $10.00733% more | 128K | 88.5%+6.5% | 95%+5.0% | 43 |
| Gemini 2.5 ProGoogle | $1.25317% more | $10.00733% more | 1M | 87.5%+5.5% | 93.5%+3.5% | 43 |
| Claude Sonnet 4.6Anthropic | $3.00900% more | $15.001150% more | 200K | 86%+4.0% | 94%+4.0% | 36 |
| Claude Sonnet 4.5Anthropic | $3.00900% more | $15.001150% more | 200K | 84.5%+2.5% | 93%+3.0% | 36 |
| Grok 4xAI | $3.00900% more | $15.001150% more | 128K | 86%+4.0% | 93%+3.0% | 33 |
| Claude Haiku 4.5Anthropic | $0.80167% more | $4.00233% more | 200K | 69.4%-12.6% | 88.1%-1.9% | 32 |
| GPT-4oOpenAI | $2.50733% more | $10.00733% more | 128K | 80.5%-1.5% | 91%+1.0% | 30 |